Draft Technical Report
GROUND WATER MODELING STUDIES AT IN
SITU LEACHING FACILITIES AND
EVALUATION OF DOSES AND RISKS TO OFF-
SITE RECEPTORS FROM CONTAMINATED
GROUND WATER
REVISION 2
U.S. Environmental Protection Agency
Office of Radiation and Indoor Air
Radiation Protection Division
1200 Pennsylvania Avenue
Washington, DC 20460
EPA-402-D-14-002
-------
Draft Technical Report
GROUND WATER MODELING STUDIES AT IN SITU LEACHING
FACILITIES AND EVALUATION OF DOSES AND RISKS TO OFF-SITE
RECEPTORS FROM CONTAMINATED GROUND WATER
REVISION 2
Contract Number EP-D-10-042
Work Assignment 2-06, Task 2
Prepared by:
S. Cohen & Associates
1608 Spring Hill Road, Suite 400
Vienna, VA 22182-2241
Prepared for:
U.S. Environmental Protection Agency
Office of Radiation and Indoor Air
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
Kenneth Czyscinski
Work Assignment Manager
December 2012
-------
Contents
List of Acronyms and Abbreviations x
Executive Summary 1
1.0 Introduction 1-1
2.0 Ground Water Modeling of In-Situ Leaching Failure Scenarios 2-1
2.1 In-Situ Leaching Process 2-1
2.2 Failures during Operations 2-3
2.3 Failures after Shutdown 2-5
2.4 Ground Water Model Development 2-6
2.4.1 Basic Aspects of Computer (Numerical) Modeling 2-7
2.4.2 Conceptual Model Development 2-8
2.4.3 Computer Code Selection 2-9
2.5 Representative ISL Facility Development and General Modeling Approach ... 2-10
2.6 Development of Scenarios 2-13
2.6.1 Spills and Leaks 2-14
2.6.2 Excursion Scenarios 2-30
3.0 Pathway Dose and Risk Conversion Factors 3-1
3.1 Pathway Dose and Risk Models 3-3
3.1.1 Ingestion of Drinking Water 3-3
3.1.2 Inadvertent Ingestion of Soil 3-3
3.1.3 Ingestion of Vegetables 3-4
3.1.4 Ingestion of Milk 3-5
3.1.5 Ingestion of Meat 3-6
3.1.6 Pathway Dose and Risk Factors 3-6
3.1.7 Implementation 3-7
3.2 Input Parameters 3-7
3.2.1 Age Groups 3-7
3.2.2 Dose and Risk Conversion Coefficients 3-9
3.2.3 Ingestion of Drinking Water 3-15
3.2.4 Inadvertent Ingestion of Soil 3-18
3.2.5 Ingestion of Vegetables 3-20
3.2.6 Ingestion of Milk 3-23
3.2.7 Ingestion of Meat 3-27
3.3 Pathway Dose and Risk Conversion Factors 3-29
3.4 Native American Exposures 3-35
3.4.1 Ingestion of Drinking Water 3-35
3.4.2 Ingestion of Vegetables 3-36
3.4.3 Ingestion of Meat 3-37
3.4.4 Exposure in Sweat Lodge 3-38
3.4.5 Native American Pathway Dose and Risk Conversion Factors 3-45
3.5 Exposure Pathways Not Analyzed in Detail 3-47
3.5.1 External Exposure to Contaminated Ground 3-48
3.5.2 Exposure to Indoor Radon 3-48
-------
3.5.3 Swimming Pool/Hot Tub Exposures 3-50
3.5.4 Exposures Due to Hydroponics and/or Aquaculture 3-50
3.5.5 Embryo and Fetus Exposure 3-51
3.5.6 Infant Consumption of Formula/Milk 3-52
4.0 Dose and Risk Assessment 4-1
4.1 Introduction 4-1
4.2 Selection of Lixiviant Concentrations and KdS 4-1
4.2.1 Lixiviant Concentration 4-1
4.2.2 Select!on of Kd Values 4-3
4.2.3 Recommendations for Kd and Radionuclide Source Term 4-6
4.3 Dose and Risk Calculations 4-7
4.3.1 Limiting Doses and Risks 4-7
4.3.2 Doses and Risks from Excursion Scenarios 4-7
4.3.3 Doses and Risks from Surface Leak Scenarios 4-13
4.4 Risks to Non-standard Receptors 4-15
4.5 Non-Radiological Risks 4-17
5.0 Summary and Conclusions 5-1
5.1 Ground Water Modeling Studies 5-1
5.2 Pathway Dose and Risk Conversion Factors 5-2
5.3 Dose/Risk Assessment for Modeled Scenarios 5-3
6.0 References 6-1
11
-------
List of Tables
Table ES-l: Calculated Total Pathway Dose Conversion Factors 4
Table ES-2: Calculated Total Pathway Risk Conversion Factors 4
Table ES-3: Typical Pathway Contributions to the Adult PDCF and PRCF 5
Table ES-4: Calculated Pathway Dose Conversion 6
Table ES-5: Excursion Scenario Maximum Doses and Risks -Mean Adult 6
Table ES-6: Excursion Scenario Doses and Risks to Various Receptors 7
Table ES-7: Surface Leak Scenario Doses and Risks 7
Table 2-1: Representative ISL Parameter Ranges 2-13
Table 2-2: Catastrophic LeakFailure Scenarios 2-19
Table 2-3: Long-Term/Low Volume Leak Scenarios 2-22
Table 2-4: Short-Term/High-Volume Leak Scenarios 2-28
Table 2-5: Series 1 - 5-Spot Injection at a Spacing of 250 feet - Receptor Well at 528
feet 2-37
Table 2-6: Series 2 - 5-Spot Injection at a Spacing of 50 feet - Receptor Well at 528
feet 2-38
Table 2-7: Series 3 - 7-Spot Injection at a Spacing of 250 feet - Receptor Well at 528
feet 2-41
Table 2-8: Series 4 - 7-Spot Injection at a Spacing of 50 feet - Receptor Well at 528
feet 2-44
Table 2-9: Series 5 - 5-Spot InjectionVPumping Rates Dependent Upon Hydraulic
Conductivity 2-46
Table 2-10: Series 6 - 5-Spot 20-Foot Thick Mined Interval 2-49
Table 2-11: Series 7 - Twenty-five 5-Spot Pumping/Injection Cells 2-52
Table 2-12: Abandoned Borehole Simulations 2-54
Table 2-13: Confining Bed Discontinuity Simulations 2-58
Table 3-1: Age Groups Used in the Analysis 3-9
Table 3-2: Radionuclide-Specific Ingestion Dose and Risk Coefficients 3-11
Table 3-3: FOR 13 Drinking Water Cancer Morbidity/Mortality Risk Ratio 3-13
Table 3-4: Risk Coefficient Uncertainty 3-14
Table 3-5: Age-Dependent Weight Normalized Water Consumption Rates 3-15
Table 3-6: Body Weight Distributions 3-16
Table 3-7: Soil + Dust Ingestion 3-19
Table 3-8: Element-Specific Input Parameters for Lognormal Distribution 3-20
Table 3-9: Age-Dependent Weight Normalized Vegetable Consumption Rates 3-22
Table 3-10: Age-Dependent Weight Normalized Milk Consumption Rates 3-26
Table 3-11: Age-Dependent Weight Normalized Meat Consumption Rates 3-28
Table 3-12: Calculated Total Pathway Dose Conversion Factors 3-29
in
-------
Table 3-13: Calculated Total Pathway Risk Conversion Factors 3-30
Table 3-14: Typical Pathway Contributions to the Adult PDCF and PRCF 3-31
Table 3-15: U-238+P Pathway Contributions to the Adult PDCF and PRCF 3-32
Table 3-16: Typical Pathway Contributions to the Teen PDCF and PRCF 3-33
Table 3-17: Typical Pathway Contributions to the Child PDCF and PRCF 3-34
Table 3-18: Typical Pathway Contributions to the Infant PDCF and PRCF 3-34
Table 3-19: Radionuclide-Specific Inhalation Dose and Risk Coefficients 3-39
Table 3-20: Radionuclide-Specific Submersion Dose and Risk Coefficients 3-40
Table 3-21: Other Parameters Used to Evaluate the Sweat Lodge Pathway 3-41
Table 3-22: Air Moisture Content with Heat Indices of 130° F and 180° F 3-44
Table 3-23: Calculated Native American Total Pathway Dose and Risk Conversion
Factors 3-46
Table 3-24: Native American Pathway Contributions to the PDCF 3-46
Table 3-25: Native American Pathway Contributions to the PRCF 3-47
Table 3-26: External Exposure to Contaminated Ground Screening Study Results 3-48
Table 3 -27: Data Used to Evaluate the PDCF for Home Radon Exposure 3-49
Table 3-28: PDCF for Home Radon Exposure 3-50
Table 3-29: Water Immersion Screening Calculation 3-50
Table 3-30: Dose Coefficients to the Offspring from Chronic Intake by the Mother 3-51
Table 3-31: PDCF to the Offspring from Chronic Intake by the Mother 3-52
Table 3-32: Dose and Risk Coefficients from Breast Milk Consumption 3-53
Table 3-33: PDCF/PRCF for Infant Milk Consumption 3-54
Table 4-1: Representative Concentrations in Uranium Alkaline ISR Lixiviants 4-1
Table 4-2: Highest Observed Concentrations in Pregnant Lixiviants based on a
Survey of Licensing Documents 4-2
Table 4-3: Estimated Range of Kd Values for Lead as a Function of Soil pH, and
Equilibrium Lead Concentrations 4-4
Table 4-4: Estimated Range of Kd Values for Uranium based on pH 4-5
Table 4-5: Radium Kd Values by Soil Type 4-5
Table 4-6: Summary of Kd Values and Lixiviant Concentrations Used in Dose and
Risk Calculations 4-6
Table 4-7: Radionuclide Well Limiting Concentrations 4-7
Table 4-8: ISL Site Hydraulic Data 4-8
Table 4-9: Excursion Scenario Maximum Doses and Risks -Mean Adult 4-11
Table 4-10: Surface Leak Scenario Doses and Risks - Mean Adult 4-14
Table 4-11: Excursion Scenario Non-standard Receptor Doses and Risksa 4-16
Table 4-12: Leak Scenario Non-standard Receptor Doses andRisksa 4-16
Table 4-13: Highest Contaminant Levels in Pregnant Lixiviant (mg/L) 4-18
Table 4-14: Scenario Maximum Contaminant Levels Versus Limit 4-18
IV
-------
Table 5-1: Typical Pathway Contributions to the Adult PDCF and PRCF 5-2
Table 5-2: Typical Pathway Contributions to the PDCF and PRCF for U-234 or U-238 5-2
Table 5-3: Radionuclide Source Term Limiting Concentrations - Receptor Well
at 528 ft 5-3
Table 5-4: Summary of Excursion Runs 5-4
Table 5-5: Number of Simulations Resulting in Various Dose Levels as Function of
Conductivity x Gradient Product 5-5
Table 5-6: Excursion Scenario Maximum Doses and Risks 5-6
Table 5-7: Summary of Dose Rate for Leak Scenarios - Adult Male Exposed to U nat
at Receptor Well 328 ft Down-gradient 5-6
Table 5-8: Comparison of Excursion Scenario Non-standard Receptor Doses and Risks
Relative to 90th Percentile Adult.
...5-7
-------
List of Figures
Figure 1-1: In-situ Uranium Recovery -Process Flow Diagram 1-2
Figure 2-1: Idealized Schematic Cross Section to Illustrate Ore-Zone Geology and
Lixiviant Migration from an Injection Well to a Production Well (NRC
2009b) 2-1
Figure 2-2: Schematic Diagram of a Wellfield Showing Typical Injection/Production
Well Patterns, Monitoring Wells, Manifold Buildings, and Pipelines (NRC
2009b) 2-2
Figure 2-3: Plan View of the Model Grid for Leak Scenarios 2-15
Figure 2-4: Cross Sectional View of the Model Grid for Leak Scenarios 2-16
Figure 2-5: Relative Concentrations for L10 at 140 Days 2-23
Figure 2-6: Relative Concentrations for LI 1 at 140 Days 2-24
Figure 2-7: Relative Concentrations for L13 at 877 Days 2-25
Figure 2-8: Relative Concentrations for L15 at 877 Days 2-26
Figure 2-9: Maximum Relative Concentrations versus Time for All Leak Simulations 2-29
Figure 2-10: Plan View of the Model Grid for Excursion Scenarios (Series 1 through 7) .... 2-31
Figure 2-11: Cross Section View of the Model Grid for Excursion Scenarios
(Series 1 through 7) 2-32
Figure 2-12: Plan View of PumpingMnjection Well Configurations and Receptor
Locations for Excursion Scenarios (Series 1, 2, 5 and 6) 2-33
Figure 2-13: Plan View of PumpingMnjection Well Configurations and Receptor
Locations for Excursion Scenarios (Series 3 and 4) 2-34
Figure 2-14: Example of Breakthrough Curves at Receptor Locations 2-35
Figure 2-15: Plan View of the Twenty-five 5-spot Pumping/Injection Well
Configurations and Receptor Locations for Excursion Scenario (Series 7) 2-51
Figure 2-16: Plan View of the Model Grid for the Abandoned Borehole and
Discontinuous Confining Bed Excursion Simulations 2-55
Figure 2-17: Cross-Sectional View of the Model Grid for the Abandoned Borehole and
Discontinuous Confining Bed Excursion Simulations 2-56
Figure 2-18: Maximum Relative Concentrations versus Time for All Excursion
Simulations 2-60
Figure 2-19: Effect of Well Spacing at Constant Hydraulic Conductivity of 100 ft/day
and Constant Hydraulic Gradient of 0.001 2-61
Figure 2-20: Effect of Well Spacing at Constant Hydraulic Conductivity of 100 ft/day
and Hydraulic Gradients of 0.1(Runs la, 2a) or 0.001 (Runs Ic, 2c) 2-61
Figure 2-21: Effect of Well Spacing at Constant Hydraulic Conductivity of 100 ft/day
and Hydraulic Gradients of 0.01(Runs Ib, 2b) and 0.001(Runs Ic, 2c) 2-62
Figure 2-22: Effect of Well Spacing on Concentration as a Function of Hydraulic
Conductivity at a Constant Hydraulic Gradient of 0.1 2-62
VI
-------
Figure 2-23: Effect of Well Spacing on Concentration as a Function of Hydraulic
Conductivity at a Constant Hydraulic Gradient of 0.01 2-63
Figure 2-24: Relative Concentrations as a Function of Hydraulic Conductivities
and Gradients 2-63
Figure 2-25: Relative Concentrations as a Function of Hydraulic Gradient
(Run Id - 0.1, Run Ig - 0.01) and Hydraulic Conductivity of 1 ft/day 2-64
Figure 2-26: Relative Concentrations as a Function of Hydraulic Gradient
(Run If- 0.1, Run li - 0.01) and Hydraulic Conductivity of 100 ft/day 2-64
Figure 2-27: Correlation of Relative Concentration to Pumping Rate 2-65
Figure 2-28: Correlation of Relative Concentration to Hydraulic Gradient 2-66
Figure 2-29: Correlation of Relative Concentration to Hydraulic Conductivity (ft/day) 2-67
Figure 2-30: Correlation of Relative Concentration to Hydraulic Gradient times
Hydraulic Conductivity 2-68
Figure 2-31: Correlation of Relative Concentration to Well Spacing 2-69
Figure 2-32: Correlation of Relative Concentration to Pumping Array 2-70
Figure 2-33: Correlation of Relative Concentration to Model Layer 2-71
Figure 3-1: Exposure Pathways Analyzed 3-2
Figure 3-2: Uranium Decay Series 3-10
Figure 3-3: Distribution of Age-Dependent Weight Normalized Water Consumption
Rates 3-16
Figure 3-4: Distribution of Age-Dependent Water Consumption Rates Compared to
Regulatory Guide 1.109 Recommended Maximum Rates 3-17
Figure 3-5: Distribution of Age-Dependent Water Consumption Rates Compared to the
RESRAD Distribution 3-18
Figure 3-6: Distribution of Irrigation Water Application Rate 3-19
Figure 3-7: Distribution of Vegetable Yields 3-22
Figure 3-8: Distribution of Age-Dependent Weight Normalized
Vegetable Consumption Rates 3-23
Figure 3-9: Distribution of Fodder (Dry) Yield 3-25
Figure 3-10: Distribution of Age-Dependent Weight Normalized Milk Consumption
Rates 3-26
Figure 3-11: Distribution of Age-Dependent Weight Normalized Meat Consumption
Rates 3-29
Figure 3-12: Cumulative Distributions of Pathway Contribution to the Total Adult U-
238+PPCDF/PRCF 3-33
Figure 3-13: Native American Drinking Water Rate Within the
Exposure Factors Handbook Distribution 3-36
Figure 3-14: Native American Vegetable Consumption Within the
Exposure Factors Handbook Distribution 3-37
Figure 3-15: Native American Meat Consumption Within the
Exposure Factors Handbook Distribution 3-38
vn
-------
Figure 4-1: Cumulative Distribution of ISL Site Hydraulic Data 4-9
Figure 4-2: Excursion Scenario 6E Peak Dose Arrival Time, Kd = 0.4 ml/g 4-10
Figure 4-3: Excursion Scenario Dose Results versus Time 4-10
Figure 4-4: Excursion Scenario Dose Results versus Hydraulic Data 4-11
Figure 4-5: Excursion Scenario Risk Results versus Time 4-12
Figure 4-6: Excursion Scenario 6E Peak Dose Arrival Time, Kd = 4.0 ml/g 4-13
Figure 4-7: Surface Leak Scenario Risk Results versus Time 4-14
Figure 4-8: Surface Leak Scenario Risk Results versus Time 4-14
Figure 4-9: 28 Day Surface Leak Scenario Doses from Uranium 4-15
Vlll
-------
Appendices
Appendix A: Verification of Codes
Appendix B: Concentration Breakthrough Curves for Leak Scenarios Runs L-l through L-24
Appendix C: Concentration Breakthrough Curves for Excursion Runs la through 7g
Appendix D: Concentration Breakthrough Curves for Abandoned Boreholes and Discontinuous
Confining Unit Runs AB-R1, AB-R2, AB-R3, CBD-R1, and CBD-R2
IX
-------
LIST OF ACRONYMS AND ABBREVIATIONS
AMG algebraic multi-grid
ATSDR Agency for Toxic Substances and Disease Registry
Bq/kg Becquerel per kilogram
CC Change Control
CCA Compliance Certification Application
CFR Code of Federal Regulations
cm3/g cubic centimeter per gram
CNWRA Center for Nuclear Waste Regulatory Analyses
COPC contaminant of potential concern
CRA Compliance Recertification Application
CTUTR Confederated Tribes of the Umatilla Indian Reservation
CVS Concurrent Versions System
DD Design Document
DDREF dose and dose-rate effectiveness factor
DF Decontamination Factor
DFSL Sweat Lodge Evaporation Decontamination Factor
DOE U.S. Department of Energy
DW drinking water
EFH Exposure Factors Handbook
EPA U.S. Environmental Protection Agency
FOR 13 Federal Guidance Report No. 13
ft/day foot per day
g/cm3 gram per cubic centimeter
g/day gram per day
GCGCD Goliad County Groundwater Conservation District
GM geometric mean
gpm gallons per minute
GSD geometric standard deviation
I&C Installation & Checkout
ICRP International Commission on Radiological Protection
ID Implementation Document
ISCORS Interagency Steering Committee on Radiation Standards
ISL in-situ leaching
-------
ISR in-situ recovery
Kd distribution coefficient
r\
kg/m kilogram per square meter
kg/yr kilogram per year
L liter
LCF latent cancer fatality
LET linear energy transfer
LMG Linked algebraic Multi-Grid solver
m meter
m3 cubic meter
MSL Sweat Lodge Air Moisture Content
mg/L milligram per liter
ml/g milliliter per gram
mrem millirem
NAS National Academy of Sciences
NCRP National Council on Radiation Protection and Measurements
NRC U.S. Nuclear Regulatory Commission
NWS National Weather Service
PABC Performance Assessment Baseline Calculation
PCG Preconditioned Conjugate Gradient
pCi picocurie
PDCF pathway dose conversion factor
pH measure of the acidity or alkalinity of a solution
PHA public health assessment
PRCF pathway risk conversion factor
psig pounds per square inch, gauge
QA Quality Assurance
RBE relative biological effectiveness
RCRA Resource Conservation and Recovery Act
RD Requirements Document
Rf retardation factor
SAB Science Advisory Board (EPA)
SC&A S. Cohen and Associates
SCMS Software Configuration Management Plan
XI
-------
SIP Strongly Implicit Procedure
SPR Software Problem Report
Sv si evert
UCL upper control limits
UIC underground injection control
UM User's Manual
UMTRCA Uranium Mill Tailings Radiation Control Act
U nat natural uranium
USCB U.S. Census Bureau
USDA U.S. Department of Agriculture
USGS U.S. Geological Survey
VD Validation Document
VMS Versions Management System
VVP Verification and Validation Plan
WA work assignment
WDEQ Wyoming Department of Environmental Quality
WIPP Waste Isolation Pilot Plant
WSDOH Washington State Department of Health
|iCi/ml microcurie per milliliter
xn
-------
EXECUTIVE SUMMARY
EPA is proposing revisions to 40 CFR 192 to create a new subsection dealing specifically with
in-situ mining of uranium. Since 40 CFR 192 was promulgated in 1983, there has been a shift in
uranium recovery from conventional milling to in-situ leaching (ISL). In the ISL process,1
chemical solutions are pumped underground through an array of wells into the ore body, where
the uranium is dissolved in place. The uranium-rich solutions are pumped to the surface, where
the uranium is extracted. The solutions are then chemically refortified and pumped back into the
ore body to recover additional uranium. Several mechanisms can be postulated by which
contaminants can leak from the ISL site and migrate to off-site wells. This report describes the
ground water modeling studies and the calculated doses and risks to receptors residing down-
gradient from a hypothetical ISL site who obtain water for drinking and agricultural purposes
from a contaminated well on their property.
The steps involved in modeling doses and risks to receptors from contaminated ground water
contaminated with radionuclides are as follows:
(1) Based on an assumed initial unit source concentration d for the /* radionuclide (1 mg/L),
calculate the relative concentration in the ground water at a down-gradient receptor well
(Cp). Convert to units of pCi/m3 based on specific activities.
(2) Develop pathway dose and risk conversion factors for all components of the ingestion
pathway, such as drinking water. For example, based on an annual drinking water
consumption Ingwat, (m3/yr) and a dose conversion factor DCing (mrem per pCi) from
EPA's Federal Guidance Report 13, the pathway dose conversion factor PDCFwat
(mrem/yr per pCi/m3) is:
PCDFwat=Ingwat * DCing
(3) Multiply the relative down-gradient receptor well concentration (Q?) by the ratio of the
actual source concentration, Cs, to the unit source concentration (Cs/Ci) to obtain the
actual down-gradient concentration, Cw- Then multiply Cw by the pathway dose
conversion factor to obtain the annual dose, Ewat (mrem/yr) to an individual from
drinking contaminated water from the receptor well:
Ewat = CwxPDCFwat
The ground water modeling studies used the computer codes MODFLOW-2000 and MT3D. For
all ground water simulations, an arbitrary source term concentration of 1 mg/L was assumed and
concentrations at down-gradient receptor wells were calculated relative to this source term. The
simulations were based on a hypotheticqal ISL site rather than a specific site and input
parameters were selected according
1 In-situ leaching (ISL) is also referred to as in-situ recovery (ISR).
ES-1
-------
Scenarios evaluated included excursions from wellfields which remain contained in the ore-
bearing aquifer, catastrophic surface leaks such as that which might occur from a break in
process piping, and slow surface leaks. For the surface leak scenarios, the leakage was assumed
to reach an aquifer overlying the ore-bearing aquifer. Excursions from the ore zone to an
overlying aquifer through an abandoned borehole and or through a natural pathway (e.g., a
discontinuity in the intervening aquitard) between the ore-bearing aquifer and an overlying
aquifer were also modeled.
A total of 63 excursion simulations involving transport of contaminants within the ore-bearing
aquifer were run, of which 49 were unique simulations and the remaining 14 duplicated other
model runs. Receptor wells for the leak scenarios were assumed to be located 328 ft, 656 ft,
1,640 ft and 3,280 ft down-gradient, while the excursion scenarios set the receptor well distances
at down-gradient distances of 528 ft, 856 ft, 1,840 ft and 3,480 ft. The excursion scenarios
involved operating an injection/extraction pattern(s) for 3 years and then shutting down the
injection wells, but continuing to remove fluids from the extraction wells. The simulations were
designed to evaluate the sensitivity of various physical and hydrogeologic properties to the
relative down-gradient concentrations at receptor wells. Variables examined in the excursion
simulations included:
• Well spacings (50, 150, and 250 ft)
• Hydraulic conductivity (1, 10, and 100 ft/day)
• Hydraulic gradient (0.001, 0.01, and 0.1 ft/ft)
• Pumping pattern (5-spot, multiple 5-spot, and 7-spot)
• Injection rates (7, 50, 150, and 500 gpm)
• Ore zone thickness (20 and 70 ft)
In all scenarios, the extraction rate was 2% greater than the injection rate during operation.
It is difficult to develop general conclusions from these excursion simulations because results of
comparisons designed into the modeling runs are at times counter-intuitive. In spite of this
difficulty, some conclusions are provided below, but the reader is cautioned to read the full text
for complete understanding as to the limitations of these conclusions:
• As expected, steeper hydraulic gradients result in shorter travel times. Furthermore, since
the pumping/injection wells are altering the regional hydraulic gradients, the arrival times
are not linearly scaled.
• An increase in hydraulic conductivity causes the contaminant plume to become more
elongated and leads to lower relative peak concentrations.
• At higher regional gradients, wider well spacings provide better capture of the lixiviant.
At lower regional hydraulic gradients, however, better capture can be maintained at
smaller well spacings.
• The 7-spot well pattern results in lower relative peak concentrations for all of the runs as
compared to a 5-spot pattern.
• The effect of pumping/injection on hydraulic gradients is strongly affected by the
transmissivity (i.e., hydraulic conductivity multiplied by thickness) of the geologic units.
ES-2
-------
The lower transmissivity results in shorter times to peak arrivals at the low and high
gradients and a longer time at the medium gradient. These relationships are all related to
how the regional and localized gradients created by the pumping/injection interact to
form a capture zone. It also illustrates the complexity and need to understand the site-
specific geology and flow system, since the effects of the interactions are not always
intuitive.
• Comparison of a single 5-spot pattern with a multiple (25) 5-spot pattern shows that
relative concentrations at down-gradient receptor wells are lower with the multiple 5-spot
pattern. These results are explained by the larger capture zone that is created by the array
of pumping/extraction wells.
Transport from the mined aquifer to an overlying aquifer through an abandoned borehole that
was not properly cemented was also evaluated. This excursion scenario can result in significant
down-gradient leakage. This emphasizes the need to carefully cement, test and inspect
abandoned boreholes to insure their integrity.
As the first step in the dose/risk assessment, probabilistic pathway dose and risk conversion
factors (PDCFs and PRCFs) were developed for the ingestion exposure pathway. For most of the
scenarios considered here, ingestion is the only significant pathway. PDCFs/PRCFs relate the
dose/risk received by an individual who utilizes the contaminated well water to the radionuclide
concentration in the well water (e.g., millirem/year per picocurie/m3). Radionuclide specific
pathway dose/risk conversion factors were calculated for U-234, U-238, Th-230 Ra-226, and
Pb-210 based on conversion factors from FGR-13 and its supporting documents. Pathways
evaluated included ingestion of milk, meat, water, contaminated soil, and vegetables. The basic
mathematical models used to calculate the dose and risk from the ingestion pathways for this
analysis were obtained from the Nuclear Regulatory Commission's (NRC's) Regulatory Guide
1.109. Although the numerical values for many of the parameters given in Regulatory Guide
1.109 have been updated since its publication more than 30 years ago, the basic mathematical
models remain valid and form the basis for many of today's computer programs used to calculate
radiological impacts. While the Regulatory Guide 1.109 models form the basis for many of
today's computer programs, those computer programs often used more refined models to better
reflect reality. When appropriate, these refined models have been used in this analysis.
Probabilistic PDCFs and PRCFs were generated for four age groups (Infant, Child, Teen and
Adult) using Excel spreadsheets and Crystal Ball to execute Monte Carlo calculations. Table
ES-1 and Table ES-2 summarize these factors. Mean pathway dose conversion factors were
lowest for adults and increased for younger people.
ES-3
-------
Table ES-1: Calculated Total Pathway Dose Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
Mean PDCF (mrem/yr / pCi/m3)
6.28E-03
1.30E-02
6.02E-04
1.61E-04
1.55E-04
7.33E-03
1.98E-02
3.62E-04
1.53E-04
1.47E-04
1.10E-02
2.62E-02
2.89E-04
1.33E-04
1.40E-04
3.89E-02
7.06E-02
4.27E-03
4.19E-04
4.24E-04
Median PDCF (mrem/yr / pCi/m3)
5.16E-03
1.01E-02
5.30E-04
1.40E-04
1.35E-04
5.17E-03
1.39E-02
2.76E-04
1.15E-04
1.10E-04
6.91E-03
1.65E-02
2.54E-04
1.01E-04
1.06E-04
3.77E-02
5.76E-02
4.37E-03
4.11E-04
4.16E-04
90th Percentile PDCF (mrem/yr / pCi/m3)
1.06E-02
2.23E-02
1.09E-03
2.77E-04
2.68E-04
1.31E-02
3.55E-02
7.43E-04
2.75E-04
2.64E-04
1.40E-02
4.28E-02
5.09E-04
1.93E-04
2.02E-04
6.31E-02
1.14E-01
7.28E-03
6.97E-04
7.06E-04
Table ES-2: Calculated Total Pathway Risk Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
Mean PRCF (LCF/yr / pCi/m3)
1.12E-09
2.25E-09
3.07E-11
2.21E-11
2.28E-11
1.62E-09
3.83E-09
4.52E-11
4.22E-11
4.91E-11
2.90E-09
7.25E-09
4.69E-11
6.92E-11
9.00E-11
4.33E-09
7.60E-09
1.31E-10
1.13E-10
1.54E-10
Median PRCF (LCF/yr / pCi/m3)
9.24E-10
1.73E-09
2.69E-11
1.91E-11
1.98E-11
1.14E-09
2.69E-09
3.44E-11
3.16E-11
3.67E-11
1.82E-09
4.56E-09
4.13E-11
5.23E-11
6.80E-11
4.20E-09
6.23E-09
1.34E-10
1.11E-10
1.51E-10
ES-4
-------
Table ES-2: Calculated Total Pathway Risk Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
90th Percentile PRCF (LCF/yr / pCi/m3)
1.89E-09
3.86E-09
5.60E-11
3.79E-11
3.92E-11
2.88E-09
6.88E-09
9.27E-11
7.58E-11
8.80E-11
3.69E-09
1.19E-08
8.27E-11
l.OOE-10
1.30E-10
7.00E-09
1.23E-08
2.22E-10
1.89E-10
2.56E-10
LCF = latent cancer fatality
Comparing Table ES-1 to Table ES-2 shows that the dose-to-risk relationship varies by a little
over an order of magnitude, from about 3 x 10"8 to 6* 10"7 latent cancer fatality per millirem,
depending on the radionuclide and age group.
The relative contributions for the modeled ingestion pathways for an adult are summarized in
Table ES-3.
Table ES-3: Typical Pathway Contributions to the Adult PDCF and PRCF
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210+P
78.3%
0.1%
13.5%
2.3%
0.3%
1.6%
0.3%
0.1%
0.3%
2.2%
0.5%
0.5%
Ra-226+P
51.2%
0.2%
8.8%
26.5%
0.2%
1.3%
1.1%
0.3%
0.7%
4.8%
4.1%
1.0%
Th-230
82.8%
0.3%
14.3%
2.1%
0.0%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.2%
U-234
76.2%
0.3%
13.1%
2.9%
0.3%
1.5%
0.3%
0.3%
0.4%
2.8%
0.6%
1.2%
U-238+P
76.2%
0.3%
13.1%
2.9%
0.3%
1.5%
0.3%
0.3%
0.4%
2.8%
0.6%
1.2%
The water ingestion pathway dominates for all of the radionuclides.
Deterministic PDCFs and PRCFs were also developed for Native Americans whose lifestyle was
significantly different from the standard adult receptor. Insufficient data on intake parameters
were available to do these analyses probabilistically. In addition to the ingestion pathways
evaluated for the standard receptors, exposures from sweat lodge rituals were included. Added
pathways involved submersion in a steam vapor cloud and inhalation of contaminated steam.
Results for PDCFs for mean adults and Native Americans are compared in Table ES-4. The
PDCFs for Native Americans are about 2 to 3 times higher than the mean for standard adult
receptors for Ra-226+P, Th-230, U-234, and U-238+P. For Pb-210+P, the PDCF for the mean
adult is higher than for the Native American.
ES-5
-------
Table ES-4: Calculated Pathway Dose Conversion
Factors for Native Americans and Adults (mean)
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Native American PDCF
(mrem/yr / pCi/m3)
5.10E-03
2.68E-02
1.94E-03
4.14E-04
3.70E-04
Adult (mean) PDCF
(mrem/yr / pCi/m3)
6.28E-03
1.30E-02
6.02E-04
1.61E-04
1.55E-04
Doses and risks were calculated for many of the excursion simulations within the ore-bearing
aquifer using expected concentrations in the leaking lixiviant. In some cases, the excursion
simulations used combinations of hydraulic parameters that were outside the ranges for operating
ISL sites. This was done to evaluate the sensitivity of various ground water modeling parameters
to variations in model output. However, dose and risk assessments were limited to parameter
ranges expected at operating facilities. Based on available site information, a cumulative
distribution function of hydraulic gradient x hydraulic conductivity (as a surrogate for ground
water velocity) was developed. From this distribution function, it was determined that all values
of the conductivity/gradient product were <0.13 ft/day. Of the 29 unique excursion simulations
that met the conductivity x gradient cutoff of 0.13 ft/day, the dose from uranium was >15
mrem/yr in 17 simulations. The highest estimated dose from uranium at a receptor well 528 ft
down-gradient was 10,072 mrem/yr, while the lowest estimated dose was 1.69E-12 mrem/yr.
The annual uranium dose at 528 ft down-gradient was less than 15 mrem for all of the scenarios
with a hydraulic conductivity of 1 ft/day and a hydraulic gradient of 0.001.
Table ES-5 shows the maximum calculated dose for various radionuclides from all 37 excursion
scenarios that were analyzed for the mean adult. Clearly, uranium and Ra-226 (+ progeny) are
the significant contributors to dose and risk. The contribution from Th-230 is one to two orders
of magnitude lower.
Table ES-5: Excursion Scenario Maximum Doses and Risks -
Mean Adult
Nuclide
Unat
Th-230
Ra-226+P
Dose (mrem/yr)
l.OE+04
2.4E+02
2.8E+04
Risk (LCF/yr)
1.4E-03
1.2E-05
4.8E-03
Table ES-6 summarizes doses and risks to various receptors from excursion scenarios. From this
table, it is apparent that the Mean Infant is the recipient of the largest calculated doses and risks.
The Mean Infant doses and risks are about a factor of three to eight times larger than the standard
receptor doses and risks. The dose and risk ratios of the other non-standard receptors to the
standard receptor are less than for the Mean Infant. For example, the Native American has a
calculated uranium risk that is 3.1 times greater than the standard receptor's mean uranium risk.
The maximum doses and risks in Table ES-6 are for particular combinations of wellfield design
parameters and hydrogeologic parameters. However, as described above, various combinations
ES-6
-------
of wellfield design and hydrogeologic parameters based on actual ISL facilities can result in
annual doses of less than 15 mrem.
Table ES-6: Excursion Scenario Doses and Risks to
Various Receptors
Receptor
90th Percentile Adult
Mean Teenager
Mean Child
Mean Infant
Native American
Nuclide
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Maximum
Dose
(mrem/yr)
1.7E+04
4.4E+02
4.7E+04
9.6E+03
1.5E+02
4.2E+04
8.7E+03
1.2E+02
5.6E+04
2.7E+04
1.7E+03
1.5E+05
2.5E+04
7.8E+02
5.7E+04
Risk
(LCF/yr)
2.5E-03
2.2E-05
8.2E-03
2.9E-03
1.8E-05
8.2E-03
5.1E-03
1.9E-05
1.5E-02
8.5E-03
5.3E-05
1.6E-02
4.3E-03
4.0E-05
l.OE-02
a - Doses for Ra-226 include progeny
Three surface leakage scenarios were evaluated: (1) catastrophic spills ranging from 100,000 to
200,000 gallons, (2) a slow leak of 1 to 2 gpm for a period of 3 years, and (3) leaks varying from
1 to 40 gpm over a 28-day period. The highest doses were incurred for scenarios involving a
slow leak over a 3-year period, while the lowest doses resulted from a 1 gpm surface leak over a
28-day period. For all scenarios, the annual doses to a mean adult from U nat were greater than
15 mrem. Results are summarized in Table ES-7.
Table ES-7: Surface Leak Scenario Doses and Risks
Unat
Th-230
Ra-226 + P
Dose (mrem/yr)
Minimum
3.2E+01
7.6E-01
8.7E+01
Maximum
1.7E+03
4.0E+01
4.6E+03
Risk (LCF/yr)
Minimum
4.5E-06
3.9E-08
1.5E-05
Maximum
2.4E-04
2.0E-06
7.9E-04
The scenarios examined focus on failures of the ISR operations that are possible but unlikely if
the operations are carefully monitored by the operators and the regulatory authorities. These
failure scenarios contain some conservative assumptions that are typical of risk assessments and
may not apply in any specific situation because of population distribution differences and the
magnitude and duration of the exposure scenarios examined versus an actual occurrence. Results
of these calculations point to the need for a rigorous regulatory regime applied to ISR operations.
ES-7
-------
Rigorous monitoring of the ISR operations would minimize the potential for these failure
scenarios and minimize exposures if they were to happen.
ES-8
-------
1.0 INTRODUCTION
In 1983, the U.S. Environmental Protection Agency (EPA) promulgated regulations at 40 CFR
Part 192 - Health and Environmental Protection Standards for Uranium and Thorium Mill
Tailings in response to the statutory requirements of the Uranium Mill Tailings Radiation
Control Act (UMTRCA) of 1978 and the Atomic Energy Act of 1954 as amended. At the time
40 CFR 192 was promulgated, uranium recovery from ore was based almost exclusively on the
conventional milling process, where a few pounds of uranium were recovered for each ton of ore
mined and processed (milled). The residues from the milling process (the tailings) were
accumulated in large piles on the surface at the mill site. Concern that these tailings piles would
be a continuing source of radiation exposure unless properly reclaimed was the driving force
behind the passage of UMTRCA. Virtually no attention was directed to other uranium recovery
operations, such as heap leaching and in-situ leaching, since at that time, the major
environmental risk was perceived to come from the conventional uranium mill tailings piles.
EPA's Office of Air and Radiation (ORIA) is currently reviewing 40 CFR Part 192 to determine
what, if any, revisions are needed to the regulations to bring them up to date. In support of
EPA's effort, ORIA requested SC&A, Inc. (SC&A) to perform a series of studies and analyses
that evaluate the potential impacts to individuals living near an operating ISL uranium recovery
facility,
The regulations under review by EPA establish standards for protection of the public health,
safety, and environment from radiological and non-radiological hazards associated with uranium
and thorium ore processing and their associated wastes. The cross-media standards apply to
pollution emissions and site restoration. The existing 40 CFR Part 192 is utilized by the U.S.
Nuclear Regulatory Commission (NRC) and its Agreement States,2 and the U.S. Department of
Energy (DOE) in their oversight of uranium extraction facility licensing, operations, sites, and
wastes. UMTRCA requires EPA to develop health and environmental standards for both Title I
inactive mill sites administered by the DOE and Title II and future NRC-licensed sites. For
future NRC-licensed sites, the standards shall be "... consistent with the standards required under
subtitle C of the Solid Waste Disposal Act,3 as amended, which are applicable to such hazards." 4
Since 40 CFR 192 was promulgated, there has been a shift in uranium recovery from
conventional milling to in-situ leaching (ISL) where, in a sense, a portion of the milling process
is conducted underground within the ore body. In the ISL process,5 chemical solutions are
pumped underground through an array of wells into the ore body, where the uranium is dissolved
in place. A process flow diagram for a ISL facility is shown in Figure 1-1. The uranium-rich
solutions are pumped to the surface, where the uranium is extracted. The solutions are then
chemically refortified and pumped back into the ore body to recover additional uranium.
2 There are currently 37 Agreement States that are responsible through the state radiation control directors and
staff, under authority of Section 274 of the Atomic Energy Act of 1954, as amended, to regulate certain uses of
radioactive materials within the state.
3 Now known as the Resource Conservation and Recovery Act (RCRA).
4 U.S.C. Title 42, Chapter 23, Section 2022.
5 In-situ leaching (ISL) is also referred to as in-situ recovery (ISR).
1-1
-------
Resin Transfer
{Cleaned resin stripped
of uranium)
Source: http: v/w\v.tceq.texas.gov assets public permitting iaclinsituin_situ diagram.jpg
Figure 1-1: In-situ Uranium Recovery - Process Flow Diagram
ORIA requested advice from the EPA Science Advisory Board (SAB) related to the Agency's
review of 40 CFR Part 192 with regard to ISL facilities. SC&A provided a separate report
summarizing relevant background information and statistical modeling approaches to assist EPA
in defining the technical issues to be considered by the SAB (SC&A 201 la). Additionally,
SC&A performed several other technical ISL studies in support of ORIA's review of Part 192.
For example, under Task 4 of Work Assignment (WA) 1-11 (Contract EP-D-10-042), SC&A
was required to collect information pertinent to site characteristics of existing heap and in-situ
leaching operations, including ground water quality and chemistry before and after leaching.
This information was summarized in a report entitled Modeling of Heap Leach andln-Situ
Leaching Operations (SC&A 201 Ib). This report was revised under Task 2 of WA 2-06 (SC&A
201 Ic) (Contract EP-D-10-042). In addition, under WA 4-17, Task 6 (Contract No. EP-D-05-
002), SC&A prepared a database of ground water information as described in Database
Summary Analysis Report: Uranium Mills andln-Situ Leach Facilities (SC&A 2008). These
reports provided background information used to develop the current report.
This report presents the results of ground water modeling studies on possible releases from
hypothetical ISL faculties based on a range of release scenarios. The studies were designed to
demonstrate whether or under what conditions leakage of various types from an ISL site to a
down-gradient receptor well can result in hazardous situations. The studies do not address
1-2
-------
specific sites. The modeling studies considered the effects of a range of site and hydrologic
variables on releases to a down-gradient receptor well. The intent was to systematically examine
how changing these parameters alters the concentrations of various potentially hazardous
constituents (mainly radionuclides) at down-gradient well relative to the concentration in the ore
zone. As such, the ground water modeling studies did not consider retardation effects, such as
radiological decay or other removal mechanisms. The relative concentrations at the receptor
wells were subsequently converted to specific concentrations by taking into account the
concentrations of specific elements at the point of release (based on reported values from actual
ISL sites). Retardation was considered only to the extent that it staggered the arrival times of the
different radionuclides at the receptor well. These specific down-gradient concentrations were
then used to calculate doses and risks using a stochastic, ingestion-based biosphere model.
Unlike the ground water modeling, the dose and risk modeling was limited to those combinations
of hydraulic parameters that are representative of actual ISL sites. Ingestion of well water and
products contaminated by well water are typically the dominant exposure pathways.
This report is not intended to provide a template as to how ground water modeling studies should
be conducted at specific sites to support specific regulatory requirements, but rather its purpose
is to scope the magnitude of potential problems that may be caused by mining solutions escaping
from ISL wellfields. As such, in some instances assumptions were made that might result is
lower radionuclide concentrations at the receptor well and/or lower doses/risks. This approach
was acceptable here, since higher concentrations and/or doses/risks would only serve to make the
potential problem worse. This is the converse to a typical site specific ground water modeling
study, which would tend to make assumptions that result in maximizing the receptor well
concentrations and/or doses/risks.
This report is divided into six chapters. Following this introductory chapter, Chapter 2 describes
the ground water modeling studies, which employed the computer codes MODFLOW-2000 and
MT3D. Scenarios evaluated included excursions from wellfields, which remain in the ore-
bearing aquifer; catastrophic surface leaks, such as those which might occur from a break in
process liquor piping; and slow surface leaks. For the surface leak scenarios, the leakage was
assumed to reach an aquifer overlying the ore-bearing aquifer. Leaks from the ore zone to an
overlying aquifer through an abandoned borehole and or through a natural pathway between the
ore-bearing aquifer and an overlying aquifer were also modeled. One of the key objectives of the
ground water modeling studies was to evaluate the range of possible releases that might occur
under various hydrologic regimes. For example, the impact of hydraulic conductivity and
hydraulic gradient were each examined over a range covering two orders of magnitude. While
this approach was useful in delineating the sensitivity of each variable, unrealistic combinations
of parameters could result. For example, in studying the effect of hydraulic conductivity, the
procedure is to select a reasonable value of hydraulic gradient (based on reported data) and vary
the conductivity over the range selected for evaluation. In this way one obtains a clear picture of
the effect of releases as a function of hydraulic conductivity at a fixed gradient. While the
generic effects are clearly delineated, the combinations of conductivity and gradient in some
simulations may exceed actual site values. This possibility is addressed in the risk assessments
by limiting the combinations of hydraulic gradient and hydraulic conductivity to those reported
for actual ISL sites. The ground water modeling assumed a constant source term of 1 mg/L and
did not consider radionuclide-specific retardation.
1-3
-------
In Chapter 3 of this report, we describe the development of probabilistic pathway dose and risk
conversion factors (PDCFs and PRCFs). Radionuclide-specific dose and risk conversion factors
were based on Federal Guidance Report 13 (Eckerman et al. 1999) and information supporting
that document. The PDCFs and PRCFs were developed for standard receptors and for non-
standard receptors (specifically Native Americans and young children). As noted above, the
ingestion pathways are the primary exposure pathways considered, except for Native Americans,
where exposure during sweat lodge rituals is also included as an exposure pathway. Discussion
of other exposure pathways is provided to indicate that their contribution to total radiation
exposure is minimal.
In Chapter 4, doses and risks for various scenarios are developed, based on: (1) source terms for
ISL sites reported in the literature, (2) documented values of distribution coefficients (Kjs), (3)
ground water dispersion, as documented in Chapter 2, and (4) the PDCFs and PRCFs developed
in Chapter 3. Since the radionuclides of interest have long half-lives (i.e., U-234, U-238, Th-230,
and Ra-226), ground water retardation delays the time to peak dose, but not its magnitude. Doses
and risks are calculated for both standard and non-standard receptors. The standard receptor was
assumed to be an adult individual who represents the 50* percentile dose/risk. Non-standard
receptors include adult individuals at the 90th percentile, teenagers, children, infants, and Native
Americans.
Summary and conclusions are presented in Chapter 5, followed by the last chapter which
provides a list of references used to develop this report.
1-4
-------
2.0 GROUND WATER MODELING OF IN-SITU LEACHING FAILURE
SCENARIOS
2.1 In-Situ Leaching Process
During ISL operations, chemicals such as sodium carbonate/bicarbonate, ammonia, sulfuric acid,
gaseous oxygen, and hydrogen peroxide are added to the ground water to produce a concentrated
oxygen-rich leaching solution called the lixiviant. The lixiviant is injected into the production
zone to mobilize (dissolve) uranium from the underground formation, and this mobilized
uranium is pumped back to the surface for extraction at a processing plant (Figure 2-1).
Injection Well — ^
/
<^M^
\ :
\
\
i
__j
,
i
V
^x
\
\
^
" 1
I
1
I
1_—
_.
*~
/*,
£-
\
J+-
• —
TZ^MMmft^Mzmzm
Potentio metric Surface (Exaggerated)
^- {
:^_^
' *"*^I"
- Less Permeable Strata • •
^ Ore Bearing Sand
.- — * * __
^^ * ^~ ~~""~ ^^"^" A
•— ~~~ — ^"A.
'cf •-•*— — '~^£z*
^^ ^ - — . — — """
^J^
— — '
—&- Production Well
^%^^^^
^,
—^-
\^~
-^ Perforations
Figure 2-1: Idealized Schematic Cross Section to Illustrate Ore-Zone Geology and
Lixiviant Migration from an Injection Well to a Production Well (NRC 2009b)
The most common injection/pumping patterns are 5- and 7-spot (NRC 2003). The shape of the
mineralized ore body and surface topography, however, may give rise to other patterns (NRC
1997). A typical 5-spot pattern contains 4 injection wells and 1 recovery well. The dimensions of
the pattern vary depending on the mineralized zone, but the injection wells are generally between
40 to 150 ft apart. In order to effectively recover the uranium and also to complete the ground
water restoration, the wells are often completed so that they can be used as either injection or
recovery wells. During mining operations, a slightly greater volume of water will be recovered
from the mineralized zone aquifer than injected in order to create a cone of depression or a flow
gradient towards the recovery wells. A typical well arrangement using 5- and 7-spot patterns is
shown in Figure 2-2.
2-1
-------
Recovery Trunkline
Injection Trunkline Wellfield Building
' A
Oxygen
5-SpoI Pattern
Injection Well
(Located at Each
Grid Intersection)
Recovery Well
(Located at Each
Grid Center)
Patterns Repeat Through Wellfield
O
o
7-Spot Pattern
• Injector Recovery Wells
A Ore Zone Monitor Wells
O Shallow Zone Monitor Wells
(One Per 4 Acres)
Groundwater Flow
Figure 2-2: Schematic Diagram of a Wellfield Showing Typical Injection/Production
Well Patterns, Monitoring Wells, Manifold Buildings, and Pipelines (NRC 2009b)
Ore body size and geometry will also influence the number of wells in a wellfield. For example,
at the Crow Butte ISL facilities in Dawes County, Nebraska, the number of injection and
production wells varied from about 190 in the first wellfield (MU-1) to about 900 in later
wellfields (MU-5 and MU-6) (NRC 1998). Three types of wells are predominant at uranium ISL
facilities:
(1) Injection wells for introducing solutions into the uranium mineralization
(2) Production wells for extracting uranium-enriched solutions
(3) Monitoring wells for assessing ongoing operations
Deep injection wells permitted by the EPA or state and approved by NRC may also be drilled for
liquid waste disposal. Injection and production wells are connected to manifolds in a nearby
header house.
Commercial-scale uranium ISL facilities usually have more than one wellfield. For example, the
Crow Butte facility in Dawes County, Nebraska, has constructed 10 wellfields since 1991 (CBR
2007). The locations and boundaries for each wellfield are adjusted as more detailed data on the
2-2
-------
subsurface stratigraphy and uranium mineralization distribution are collected during wellfield
construction.
General information on ISL facilities is included in NRC 2009b.
2.2 Failures during Operations
Potential releases that may occur during operations involve the following scenarios: (1) spills
and leaks on the surface and subsequent transport to the ground water, and (2) excursions beyond
the injection/production wells or into other non-mined geologic units above and/or below the
mined unit.
During operations, ISL operations may affect the ground water quality near the wellfields when
lixiviant moves from the production zone to beyond the boundaries of the wellfield. This
unintended spread, either horizontally or vertically, of recovery solutions beyond the production
zone is known as an excursion. An excursion can be caused by:
• Improper water balance between injection and recovery rates
• Undetected high permeability strata or geologic faults
• Improperly abandoned exploration drill holes
• Discontinuity within the confining layers
• Poor well integrity, such as a cracked well casing or leaking joints between casing
sections
• Hydrofracturing of the ore zone or surrounding units
NRC license and underground injection control (UIC) permit conditions6 require that licensees
conduct periodic tests to protect against excursions. These include, but are not limited to:
• Conducting pump tests for each wellfield prior to operations within the wellfield to
evaluate the confinement of the production horizon.
• Continued wellfield characterization to identify geologic features (e.g., thinning
confining layers, fractures, high flow zones) that might result in excursions.
• Mechanical integrity testing of each well to check for leaks or cracks in the casing. An
excursion that moves laterally from the production zone is a horizontal excursion.
Vertical excursions occur where barren or pregnant lixiviant migrates into other aquifers
above or below the production zone.
Operators must maintain ground water monitoring programs to detect both vertical and
horizontal excursions, and must have operating procedures to analyze an excursion and
determine how to remediate it. Monitoring wells are sampled at least every 2 weeks during
6 Lixiviant injection wells are classified as Type III injection wells and must meet the requirements set forth in
40 CFR 144.
2-3
-------
wellfield operations to verify that ISL solutions are contained within the operating wellfield
(NRC 2003). Geochemical excursion indicators are identified based on wellfield pre-operational
baseline water quality.
The spacing of horizontal excursion monitoring wells is based on site-specific conditions, but
typically they are spaced about 90-150 m [300-500 ft] apart and screened in the production zone
(NRC 1997 and 2003; Mackin et al. 2001; EIA 1995). The distance between monitoring wells
and the distance of monitoring wells from the wellfield are typically similar (NRC 1997 and
2006). The specific location and spacing of the monitoring wells is established on a site-by-site
basis by license condition. These are often modified according to site-specific hydrogeologic
characteristics, such as the extent of the confining layer, hydraulic gradient, and aquifer
transmissivity. Well placement may also be modified as the licensee gains experience detecting,
recovering, and remediating these excursions. NRC licenses also include requirements to
establish monitoring wells in overlying and, as appropriate, in underlying aquifers to detect
vertical excursions. Although uranium deposits are typically located in hydrogeologic units
bounded above and below by adequately confining units, the possibility of vertical contaminant
transport must be considered. Historically, these monitoring wells are more widely spaced than
those within the host aquifer, although underlying aquifer monitoring wells may not be required
under some circumstances (Mackin et al. 2001). Frequency of vertical monitoring wells at
licensed ISL facilities has been (1)1 monitoring well per 4 acres of wellfield in the first
overlying aquifer, (2) 1 monitoring well per 8 acres in each higher aquifer, and (3)1 monitoring
well per 4 to 8 acres in the underlying aquifer (Mackin et al. 2001). These monitoring wells are
typically sampled every 2 weeks during operations.
An excursion is defined to occur when two or more excursion indicators in a monitoring well
exceed their upper control limits (UCLs) (NRC 2003).7 Alternatively, since the advent of
performance-based licensing, procedures to identify excursions can be imposed through site-
specific license conditions. For example, an excursion may be defined to occur when one
excursion indicator is exceeded in a monitoring well by a certain percentage. If an excursion is
detected, the licensee takes several steps including notifying NRC and confirming the excursion
through additional and more frequent sampling (NRC 2003, Chapter 8). As described in NRC
guidance (NRC 2003, Section 5.7.8.3), licensees typically recover from horizontal excursions by
adjusting the flow rates of the nearby injection and production wells to increase process bleed in
the area of the excursion. To address vertical excursions, licensees may adjust injection and
production flow rates in the area of the excursion and pump directly from the affected
monitoring wells or from other wells drilled for that purpose. Vertical excursions are more
difficult to recover from, persisting for years in some cases. If an excursion cannot be recovered,
"Upper control limits are concentrations for excursion indicator constituents [e.g., chloride, total dissolved
solids or bicarbonate] that provide early warning that leaching solutions are moving away from the well fields and
that groundwater outside the monitor well ring may be threatened. Excursion indicator constituents should be
parameters that are strong indicators of the in situ leach process and that are not significantly attenuated by
geochemical reactions in the aquifers. If possible, the chosen parameters should be easily analyzed to allow timely
data reporting. The upper control limit concentrations of the chosen excursion indicators should be set high enough
that false positives (false alarms from natural fluctuations in water chemistry) are not a frequent problem, but not so
high that significant ground-water quality degradation could occur by the time an excursion is identified. A
minimum of three excursion indicators should be proposed" (NRC 2003).
2-4
-------
the licensee may be required to stop injection of lixiviant into a wellfield (NRC 2003,
Section 5.7.8.3).
2.3 Failures after Shutdown
Prior to implementing post-closure monitoring, aquifer restoration activities are conducted. The
purpose of aquifer restoration is to return wellfield water quality parameters to the standards in
10 CFR 40, Appendix A, Criterion 5(B)(5) or another standard approved by NRC. Ground water
adjacent to the exempted portion of the aquifer, however, must still be protected. Prior to
wellfield operations, applicants and licensees must determine baseline ground water quality for
the production zone (NRC 2003). In their applications, applicants or licensees identify the NRC-
accepted list of constituents to be sampled. Operators may identify other constituents, or remove
constituents, as long as a basis for the constituent(s) is provided and approved by NRC. State and
other federal agencies with jurisdiction over ground water could also specify constituents, which
may or may not be included in the NRC-accepted list. In this case, the applicant would be
accountable to the subject state or federal agency for characterizing and restoring these
constituents. To determine baseline water quality conditions prior to wellfield operations,
applicants or licensees collect at least four sets of samples, spaced sufficiently in time to
establish seasonal variability, and analyze the samples for the identified constituent (NRC 2003).
An NRC-acceptable set of samples should include all wellfield perimeter monitoring wells and
all upper and lower monitoring wells. Additionally, the applicant or licensee should sample at
least one production/injection well per acre in the wellfield, or enough production/injection wells
to provide an adequate statistical population if less than one well per acre is used. NRC verifies
the accuracy of baseline water quality data by ensuring that the applicant's or licensee's
procedures include (1) acceptable sample collection methods, (2) a set of sampled parameters
that is appropriate for the site and ISL extraction method, and (3) collection of sample sets that
are sufficient to represent natural spatial and temporal variations in water quality.
After uranium recovery has ended, the ground water in the wellfield contains constituents that
were mobilized by the lixiviant. Licensees usually begin aquifer restoration in each wellfield
soon after the uranium recovery operations end (NRC 2008). Aquifer restoration criteria for the
site-specific baseline constituents are determined either on a well-by-well or wellfield-by-
wellfield basis. NRC licensees are required to return water quality parameters to the standards in
10 CFR Part 40, Appendix A, Criterion 5B(5) or to another standard approved in their NRC
license (NRC 2009a). Aquifer restoration programs typically use a combination of methods,
including (1) ground water transfer, (2) ground water sweep, (3) reverse osmosis with permeate
injection, (4) ground water recirculation, and (5) stabilization monitoring (EIA 1995; Mackin
et al. 2001; Davis and Curtis 2007). NRC allows licensees the flexibility to select the restoration
methods to be used for each wellfield (NRC 2003). The EPA or state authorized to implement
the EPA underground injection control program reviews any aquifer restoration plans for
compliance with the applicable terms and conditions of the UIC permit requirements. NRC staff
review any aquifer restoration plans for compliance with the NRC license to protect human
health, safety, and the environment.
2-5
-------
2.4 Ground Water Model Development
To investigate potential impacts caused by ISL facilities on ground water, the following activities
were performed as described in this report: (1) development of a representative ISL facility,
(2) implementation of a deterministic modeling methodology, (3) definition of release scenarios,
(4) performance of a sensitivity analysis, and (5) calculation of risks based upon potential doses
to receptors for selected simulations.
Ground water flow and contaminant transport modeling are frequently conducted at ISL facilities
to evaluate possible effects of proposed and ongoing mining. For instance, a three-dimensional
ground water flow and contaminant transport model was constructed to evaluate the ISL facility
in Goliad County, Texas (DBS&A 2007). This modeling was performed for the Goliad County
Groundwater Conservation District (GCGCD) and the major objectives of the modeling study
were to (1) evaluate the practicality of controlling injection fluid excursions from escaping
downdip, (2) evaluate the practicality of controlling injection fluid excursions from escaping
vertically into non-mined aquifer zones, and (3) determine the amount of bleed water required to
control or eliminate such excursions. ADD NRDC 2012
The modeling at the Goliad site is particularly relevant to this current task for the following
reasons: (1) several of the modeling objectives are very similar, (2) the modeling approach that
was taken is analogous to the approach undertaken here, and (3) the computer codes that were
used are identical to those used for the modeling in this report. Furthermore, as presented below,
the modeling conclusions have implications with respect to the occurrence of potential for
excursions during active mining, the placement and depths of monitoring wells, sampling
frequencies, and post-closure monitoring timeframes. As described in DBS&A 2007:
Simulation results indicate that capture of injected fluids within the mined zone
with 1 percent bleed water is feasible, although the simulation results are very
sensitive to well placement, selected injection and pumping rates, and hydraulic
conductivity of the aquifer. In some cases, an increase of even 20 feet in the well
spacing caused the simulated bleed water to increase by about 6 per cent.
In addition, the nature of the hydraulic groundwater flow field that may develop
due to mining leads to the formation of long, low-velocity travel paths in certain
areas of the injection-capture system; impacted groundwater within or near these
travel paths may not be extracted during the life of the mining operation if a
specific approach is not designed and implemented to account for these aquifer
regions. Most of the injected fluid is extracted at capture wells within a time
frame of about 3 years or less for the scenarios evaluated. At the cessation of ISL
mining, monitoring locations should be selected carefully in order to identify
potential groundwater impacts in the vicinity of these longer, low-velocity
pathways. In addition, groundwater monitoring should be continued for an
extended period of time.
The simulation results also indicate that ISL fluids can, and likely would, migrate
vertically between aquifer layers. The ISL scenario evaluated assumes that
2-6
-------
injection and pumping would occur in the uppermost layer of the Evangeline
aquifer (GoliadSand), which has a saturated thickness of approximately 105 feet
in the model. ISL fluids migrate vertically downward in the vicinity of the
injection well. Once these fluids enter the model layer below the mined layer, they
are not recaptured by the pumping wells, but rather migrate with the ambient
groundwaterflow velocity of the deeper aquifer layer.
2.4.1 Basic Aspects of Computer (Numerical) Modeling
The flow and transport modeling performed for this analysis uses finite-difference techniques,
which require that the ground water system be divided ('discretized') into finite-sized blocks or
'cells.' Each cell is assigned unique hydraulic properties depending on the available field data
and the goals for the analysis. In this way, complex features of the ground water system can be
accommodated in the model. The time represented by the modeling effort must also be divided
into discrete periods or 'time steps.' These steps must be short enough to provide an accurate
solution, but not so short that they require an excessive number of calculations to run a
simulation. The finite-difference method also requires that values for head be assigned at flow
boundaries (referred to as 'boundary conditions'), as well as for the initial time period of the
simulation (referred to as 'initial conditions'). This is a requirement for producing a unique
solution with any numerical method that depends on iteration, as does the finite-difference
method. Models were applied that simulate ground water flow and chemical transport.
The three-dimensional computer model for analyzing ground water flow generates a flow field
(array of head values) representing average conditions in the model area. The flow model is used
as the basis for the transport model.
The chemical transport model evaluates how the average flow field, along with other transport
parameters, affects chemical movement in ground water and plume development from lixiviant
sources. The chemical transport model simulates the expansion of the plume, both during the
active leaching activities time as well as the post-closure stage.
After assigning material properties and initial and boundary conditions, the finite-difference
equations for flow are solved to produce a mathematically 'approximate,' but scientifically
reliable, value of the average ground water head (potentiometric surface elevation) within each
cell. Models that use the finite-difference numerical techniques allow rapid analysis of complex,
time-dependent ground water systems.
Numerical models are operated by a computer code or program. The code is a generalized set of
steps, to which specific field conditions, such as initial and boundary conditions, are imposed.
Because computer codes are generic in nature and must be adapted to actual field conditions, a
clear understanding of the existing physical system (a conceptual model) is required.
It is important to establish why the model is being created, and to properly design the model
simulations to sufficiently address the objectives. The model development for each of the failure
scenarios followed the steps detailed below:
2-7
-------
(1) Developing a conceptual model to guide creation of model attributes
(2) Selecting an appropriate computer code(s) for the analysis
(3) Establishing the time period represented by the model and the duration of subdivisions of
this period (time steps) required for modeling
(4) Selecting a suitable model domain and determining the dimensional (horizontal and
vertical) limits of the analysis
(5) Establishing the model structure by determining the number of model layers and the grid
spacing requirements for the flow analysis
(6) Incorporating hydraulic boundaries and features, including the shape and characteristics
of constant-head boundaries, precipitation/recharge, and pumping/injection
(7) Assigning hydraulic conductivity values
(8) Specifying initial head values (ground water surface elevation)
(9) Evaluating and assigning appropriate model computational characteristics; for example,
solution method, iteration limits, and convergence criteria, to enhance model stability,
computational efficiency, and solution accuracy
(10) Evaluating the sensitivity of model results to changes in model parameters
(11) Establishing the model structure, including determining the number of model layers and
the grid spacing requirements for the transport analysis
(12) Assigning the characteristics of chemical sources (e.g., leaks, spills) consisting of
dimensions, locations, concentrations, and time dependency
(13) Assigning transport parameters, including the dispersivities and porosities
(14) Conducting chemical transport simulations and exporting the observed concentrations at
pre-specified locations
(15) Processing the data within Excel spreadsheets
2.4.2 Conceptual Model Development
The general components of the conceptual model that serve as the basis for the construction of
the ground water flow and contaminant transport models are described below. This conceptual
model summarizes the theoretical understanding of the primary conditions that affect ground
water flow and chemical transport and fate. More detailed descriptions (e.g., pumping/extraction
well configurations) are presented within discussions of the respective failure scenarios.
As contaminant plumes move down-gradient from the source area, they tend to spread laterally
and vertically, thereby lowering the average contaminant concentration as the plume expands.
The shape taken by an individual plume varies depending primarily on the nature of the geologic
materials making up the aquifer, but also secondarily on the rate of ground water flow.
In fine-grained unconsolidated sediments, such as sands and silts, plumes tend to spread out
laterally in a fan shape as they move down-gradient. This process is called dispersion. Vertical
2-8
-------
flow also occurs and is controlled by the uniformity of the sediments, as well as the vertical
hydraulic gradient. When all the aquifer materials are of essentially the same size and are well-
rounded, vertical flow can easily take place, assuming a vertical hydraulic gradient exists. Fine-
grained layers of sediments such as clays and silts in an otherwise coarse-grained aquifer prevent
or retard downward (or upward) vertical flow. Ground water flowing at a moderate to fast rate
tends to minimize both horizontal and vertical dispersion, while slower flow (normally in fine-
grained materials) allows greater dispersion. All of these processes, however, will be
complicated by the effects caused by the injection and withdrawal of ground water.
Contaminant plumes extend down-gradient from the source area over time until a steady state
condition is reached, based on the rate of contaminant flux to the ground water and the degree of
chemical degradation/sorption taking place in the aquifer. Contaminant concentrations decline as
down-gradient flow occurs, because processes such as dispersion, adsorption, and chemical
transformation are constantly taking place in the aquifer. The length of a plume will depend on
(1) how rapidly these processes work, (2) the rate of ground water flow, (3) the rate of chemical
releases to the aquifer, (4) chemical interactions between the ground water and aquifer matrix,
and (5) other environmental factors, such as temperature and the basic chemistry of the ground
water. Ultimately, even with a constant source of contamination to the aquifer, any plume will
reach a point beyond which it can no longer expand and will more or less stabilize. This
stabilization, or steady state condition, occurs when degradation and/or sorption processes in the
aquifer remove as much contaminant mass as is being released to the aquifer in the source area.
If the source of contamination is cut off, for example by pump and treat extraction wells, a
reduction in chemical concentrations will occur down-gradient of the source area and will be
especially noticeable along the axis of the plume. Over time, the reduction in plume
concentrations will be propagated farther down-gradient consistent with the hydraulic
conductivity of the aquifer. Subsequently, the plume will begin to contract in areal extent.
In the case of ISL facilities, the contaminants are mobilized by the lixiviant, which is often
injected into units that are straddled above and below by aquifers that are used as drinking
sources. In fact, at least one ISL facility injects lixiviant into the same aquifer as that used as a
nearby drinking water source (Rice 2006).
2.4.3 Computer Code Selection
The computer codes that were used for this analysis are MODFLOW-2000 and MT3D-MS.
MODFLOW-2000, the U.S. Geological Survey (USGS) finite-difference ground water flow
model, is a popular and widely used computer code (Harbaugh et al. 2000). Ground water flow
within the aquifer is simulated using a block-centered finite-difference approach. Layers can be
simulated as confined, unconfined, or a combination of confined and unconfined. Flow
associated with external stresses, such as wells, areal recharge, evapotranspiration, drains, and
streams, can also be simulated.
The modular three-dimensional (3-D) transport model referred to as MT3D-MS was originally
developed by Zheng and Wang (1999) at S.S. Papadopulos & Associates, Inc., and subsequently
documented for the Robert S. Kerr Environmental Research Laboratory of the EPA. In the past
2-9
-------
several years, various versions of the MT3D code have been commonly used in contaminant
transport modeling and remediation assessment studies. MT3D-MS does not explicitly simulate
geochemical reactions, but can be used to simulate changes in concentrations of miscible
contaminants in ground water considering advection, dispersion, diffusion and some aggregate
chemical reactions (i.e., distribution coefficient), with various types of boundary conditions and
external sources and/or sinks. The basic chemical reactions included in the model are
equilibrium-controlled or rate-limited linear or non-linear sorption, and first-order irreversible or
reversible kinetic reactions. Somewhat more sophisticated, multispecies chemical reactions can
be simulated by add-on reaction packages.
MODFLOW2000 and MT3D-MS are commonly applied at ISL facilities to evaluate how the
average flow field, together with other transport parameters, affects chemical movement in
ground water and plume development from lixiviant sources. The chemical transport model is
often used to simulate the expansion of the plume, both during the active leaching activities time
as well as the post-closure stage.
The pre- and post-processing of data input/output for these codes was performed with
Groundwater Vistas (Rumbaugh and Rumbaugh 1998).
2.5 Representative ISL Facility Development and General Modeling Approach
Prior to discussing the model input and parameter ranges, it is important that the overall goal of
the modeling be reiterated. The central question to be addressed by the modeling is whether
health-based standards may be exceeded under representative injection\extraction configurations
and realistic parameter combinations. To meet this objective, the model does not need to
simulate flow processes at the interstitial pore level (e.g., heterogeneity, dispersivity) or
geochemical processes at the molecular level (e.g., kinetics and thermodynamics). Instead
simplifying assumptions are made in the modeling to adequately capture the overall effects of
very complex processes affecting ground water flow and contaminant transport.
An important aspect in constructing the model was to ensure that the assumptions were not so
conservative as to render the results unproductive by routinely exceeding the health-based
standards. In some instances, assumptions are made that may underestimate the predicted
downgradient plume concentrations. This approach is justified because even under the
nonconservative assumptions, the health-based standards are still exceeded in some cases. It is
important to keep in mind that this modeling does not replace the need for site-specific modeling,
nor is it intended to provide a framework upon which to perform site-specific modeling.
In deciding whether to perform deterministic or stochastic modeling, several factors were
considered, including (1) the ability to meet the overall objectives, (2) difficulty in setting up and
explaining the model (e.g., treatment of correlation, defining parameter distributions),
(3) complexities of running the model (e.g., demonstrating statistical convergence); and (4) effort
required to interpret the results. After evaluating these considerations, a deterministic approach
was selected, since it could meet the objectives, is simpler to set-up and explain, and the
interpretation of the results is more straightforward.
2-10
-------
A thorough review of readily available ISL facility information indicates that there are several
important factors that need to be considered in evaluating their potential impacts on ground water
and must be conceptualized in the numerical model. These common elements provide a
framework from which the sensitivity analyses for each of the scenarios was conducted. Those
attributes that are specific to each scenario (e.g., model domain and grid spacings) will be
presented under each of the respective sections.
As part of the deterministic approach, the sensitivity of the results to the most uncertain
parameter values was assessed. These parameters include hydraulic gradients and conductivities,
leak, pumping and injection rates, and durations. All of the concentrations are calculated as
relative concentrations assuming a 1-mg/L source term and are adjusted to actual concentrations
during the dose assessment (Section 4). Ranges for the representative model input parameters are
presented in Table 2-1 and discussed below.
The regional hydraulic gradient is an estimated parameter based upon a literature review of
natural gradients under nonstressed conditions. For a set value of hydraulic conductivity and
effective porosity, a higher hydraulic gradient results in a faster ground water velocity, less
dispersion and higher doses. There may be cases where pumping at the ISL facility creates
hydraulic gradients that are greater than those estimated for the modeling. Therefore, as part of
the sensitivity analysis, the range of hydraulic gradients was increased to 0.1 ft/ft.
One of the common strategies at all of the ISL facilities is to better understand the most probable
fate and transport of uranium and other constituents during and after ISL operations. To achieve
this goal, mathematical modeling of chemical reaction kinetic equations or equilibrium
thermodynamic equations are often used to describe chemical interactions among dissolved
chemical species, the dissolution of immobile solid phases, or the formation and precipitation of
new, immobile solid phases. EPA recognizes the importance of understanding the geochemical
processes and has entered into a corporative agreement with the U.S. Geological Survey under a
Regional Applied Research Effort (RARE). The main objective of this work is to provide a
predictive model that describes the ground water flow and geochemical changes along with
longer-term transport of dissolved constituents during and after the uranium ISL mining process.
To accomplish this goal, analysis of the lithology, ore and ground water chemistry is being
characterized at the Dewey Burdock uranium project site. This site is located approximately
65 miles southeast of Rapid City, South Dakota, which is one of the areas being considered for
in-situ leaching (ISL) of uranium. The available ground water flow and transport data will be
input into MODFLOW and MT3D-MS. More quantitative reactive transport modeling will also
be conducted using PHT3D, which couples MODFLOW to PHREEQC (an advanced
thermodynamically based geochemical code).
To meet the objectives of the current ISL modeling, however, a more simplistic approach is
taken to simulate the geochemical behavior. In this method, the net effect from all of the
geochemical processes are expressed as a distribution coefficient (Kd). The distribution
coefficient is subsequently used to estimate the amount of retardation that each of the
contaminants would experience along the travel path. Since all of the key radionuclides have
long half lives, the amount of retardation does not significantly affect the peak doses. Therefore,
2-11
-------
retardation and radioactive decay are not explicitly simulated in the modeling. Arrival times (i.e.,
breakthrough curves) are corrected for retardation as part of the risk assessment presented in
Section 4.
Areal recharge impacts ground water flow and contaminant transport in several ways, including
(1) altering the horizontal and vertical hydraulic gradients, (2) affecting the geochemistry, and
(3) depressing the plume deeper into the aquifer as a function of distance. Since the analyses
were performed over a range of hydraulic gradients, the effects of recharge on the gradients are
implicitly considered. Any recharge water infiltrating to the depths being mined by the ISL
facilities will take sufficiently long that the system will approach equilibrium conditions.
Furthermore, any transient effects of recharge on geochemistry and contaminant mobility are
expected to be minimal compared to the geochemical mobilization properties of the lixivant. The
prevailing geochemical conditions within the mined unit and aquifers are assumed to result in
low sorption properties for the contaminants (Section 4.2.2).
Although areal recharge may depress the plume, these effects would not significantly affect the
peak dose and are therefore ignored. For these reasons, no areal recharge was explicitly assigned
in the modeling.
With respect to hydrostratigraphy, the major sandstone roll-front uranium deposits found in
Wyoming, South Dakota, Nebraska, New Mexico, and Texas have similar deposit!onal histories,
which have resulted in similar rock compositions. Typically, the roll-front deposits consist of
sandstones of fluvial origin that are generally interbedded between siltstones and mudstones
(NRC 2009a). Sandstones generally have moderate hydraulic conductivities, and the modeling
assumes a range from 1 to 100 ft/day (Nicot et al. 2010). To simplify the analysis, the hydraulic
conductivities in each of the simulations is assumed to be homogeneous and isotropic.
Heterogeneity and preferential pathways from the simulations are addressed in several ways.
First, the hydraulic conductivities are sampled over several orders of magnitude and should
adequately bracket the potential impacts of heterogeneities on contaminant arrival times. Second,
as discussed in greater detail below, the dispersivities have been set to the high end of the
expected range. This assumption spreads out the plume to better simulate the effects of potential
heterogeneities. Finally, although discrete features and lower dispersivities could lead to releases
greater than those predicted, the dose limits were already exceeded (see Section 4.3) without
their consideration, thereby removing the need to further evaluate their potential impacts.
Dispersivity is a geometric property of a porous medium which determines the dispersion
characteristics of the medium by relating the components of pore velocity to the dispersion
coefficient. The amount of dispersion is scale dependent and describes the degree to which the
plume spreads out and elongates along the travel path. Longitudinal dispersivity is often assigned
a value of about 10% of the travel distance. The transverse dispersivity is about 10% of the
longitudinal and the vertical dispersivity is about 10% of the transverse. The nearest receptor for
the dose assessment is assumed to be about 328 feet downgradient from the ISL facility. The
dispersivities assigned in the modeling are held constant for all simulations and are 65 ft, 6.5 ft
and 0.65 ft, for the longitudinal, transverse and vertical, respectively. Although these values are
somewhat higher than would be expected, they are reflective of potential heterogeneities that are
not captured in the homogeneous and isotropic assumptions.
2-12
-------
The effective porosity describes the amount of interconnected pore space. A constant value of
20% was used for all of the simulations. Although it is likely that the porosity and
permeability/connectivity of the aquifer will change as the uranium is extracted from the
production areas, these impacts will be greatest within the mined areas, rather than those areas
where downgradient excursions are occurring. The possibility that changes in porosity and
permeability could cause greater excursions is captured by the simulations where the high range
of hydraulic conductivities is selected.
Pumping of wells in the vicinity of the ISL facility may affect hydraulic gradients as well as the
amount of dilution. As discussed above, higher gradients have been investigated to capture
potential pumping effects. To compensate for dilution effects, the dispersivity is set to a
relatively high value and the plume is assumed to be completely mixed within the entire
thickness of the aquifer. Although this assumption may underestimate peak doses, health-based
standards are still exceeded in a number of scenarios.
Since long-term average injection/extraction rates are assumed, all of the flow simulations are
performed to steady-state solutions controlled by constant head values. Head values can change
for many reasons and over various time scales as a result of severe, short-term weather
disturbances, seasonal variations, or longer-term variations most likely due to human activities.
Long-term variations (year to decade time scales) could occur in the arid, western portion of the
United States where there could be increased competition for water resources from aquifers. It is
recognized that changes in the boundary conditions could increase or decrease the probability
and severity of excursions away from an ISL site. It is unlikely, however, that the releases would
fall outside of the range of those predicted. Furthermore, the transport analysis is conducted for a
time period that is sufficiently long to allow for contaminant concentrations to peak at the nearest
receptor well.
Table 2-1: Representative ISL Parameter Ranges
Model Input Parameter
Regional Hydraulic Gradient (ft/ft)
Recharge
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone, (milligram per L)
Hydraulic Conductivities (ft/day)
Over- and underlying aquifers
Sandstone (mined interval)
Injection/Extraction Well Spacing (ft)
Range of Injection Rates (gpm)
Operating Life of ISL Well Pattern (yrs)
Potential Range of Values
0.001-0.01
0.09-0.15
5-30
10-1000
1-100 .
0.1-10
1
10-100
1-10
40-150
20-200
1-3
References
Estimated parameter
Chowdhury and Mace 2007
Freeze and Cherry 1979
VL=0.1(plume
length)
VT=0.1(VL)
VV=0.1(VT)
Gelharetal. 1992
Normalized concentration
Nicotetal. 2010
NRC 2009b
NRC 2009b
NRC 2009b
2.6 Development of Scenarios
2-13
-------
Two scenarios are investigated that are associated with the active phase of the mining activities;
(1) spills and leaks on the surface and subsequent transport to the ground water, and
(2) excursions beyond the injection/production wells or into other non-mined geologic units
above the mined unit. The general question that the flow and transport modeling will assist in
addressing is: What concentrations of contaminants could reach nearby domestic wells and over
what timeframes? The modeling does not make any assumptions with respect to whether best
management practices are being followed, and simply relies on published data for parameter
ranges and distributions. In many cases, it is likely that a leak would be detected and remedied
before any contamination reaches nearby receptors.
2.6.1 Spills and Leaks
During ISL operations and aquifer restoration, barren and pregnant uranium-bearing process
solutions are moved through pipelines to and from the wellfield and among different surface
facilities (e.g., processing circuit, evaporation ponds). To investigate the potential effects of a
pipeline rupture or failure, the Spills and Leaks scenario investigated three general types of spills
and leaks, including (1) catastrophic failure (e.g., break of a pipe or breach of an evaporation
pond), (2) long-term/low-volume leak of an underground pipe or well transferring or
withdrawing lixiviant, and (3) short-term/high-volume leak of an underground pipe or well
transferring or withdrawing lixiviant.
The model grid is identical for all of the spills and leaks simulations and consists of a domain
that has an area of 1 square mile (Figure 2-3). It is assumed that ground water is flowing due
south and the model grid is oriented perpendicular to ground water flow. Constant head
boundaries are assigned along the northern and southern boundaries in order to simulate the
regional gradient. Boundaries on the east and west of the model domain are oriented
perpendicular to ground water flow (i.e., hydraulic divide) and, therefore, ground water does not
enter or exit the model along these boundaries (i.e., no-flow boundary conditions). This
boundary configuration makes the solutions to the flow problem more unique than would be
obtained if constant head boundaries surrounded the entire domain. The base of the model is also
represented by a no-flow boundary condition.
The aquifer is assigned a thickness of 130 ft that is discretized into 10 layers (Figure 2-4).
Horizontal grid spacing is on 10-ft centers in the vicinity of the spill, and is gradually increased
to a spacing of 250 ft at the model boundaries. The vertical grid spacing ranges from 2 ft at the
top of the model to 50 ft at the base. The smaller grid spacing in the area of the contaminant
source reduces numerical dispersion, provides better mass balance and allows for more rapid
numerical convergence of the solver.
2-14
-------
Constant Head Boundaries
(U
o
00
CM
Constant Head Boundaries
5280 feet
Receptor Locations
Figure 2-3: Plan View of the Model Grid for Leak Scenarios.
2-15
-------
South
Cross-Section Along Column 31
Spill/Leak
North Thickness (feet)
1
2
3
4
5
10
15
20
25
35
Figure 2-4: Cross Sectional View of the Model Grid for Leak Scenarios.
All fluid entering the model does so as recharge through a 100-ft2 area. The rationale for this
assumption is that leaks from even small releases are likely to spread out at least 10 ft as the fluid
migrates through the unsaturated zone.
The minerals within the vadose zone may have significantly different chemical properties than
those of the aquifer leading to different solute retardation behavior than assumed for the
underlying aquifer. In the case where there is greater sorption in the vadose zone than in the
2-16
-------
underlying aquifer, receptor dose could be significantly decreased. In some cases, the
contaminants may diffuse into the finer grained matrix, precipitate and/or be adsorbed. This
could result in the vadose zone being a potential long-term source of contaminants that could
leach into the ground water. None of these processes are explicitly simulated in the model and
flow through the unsaturated zone is assumed to be instantaneous. This assumption is made
because the focus of the modeling is on cases where the unsaturated zone does not inhibit the
migration of contaminants or act as a long-term contaminant sink.
Hydraulic conductivities are assigned homogeneous and isotropic properties. Since contaminant
concentrations at receptor locations are averaged across the entire aquifer to simulate pumping
effects, this assumption will not bias the results. Hydraulic conductivities are assigned values of
1,10 and 100 ft/day to represent the transmissive properties of the aquifer. These values are
within those ranges presented for overlying aquifers in the south Texas uranium province (Nicot
et al. 2010), and are at the high end of the range for lithologic units within the United States
where ISL mining is practical.
Potential receptors are assumed to be located at 328 ft (100 m), 656 ft (200 m), 1,640 ft (500 m)
and 3,280 ft (1,000 m). In the discussion that follows, all of the relative peak arrival times and
concentrations cited are for the potential receptor located at 328 ft down-gradient of the release.
Breakthrough curves (i.e., concentrations as a function of time and distance) for all of the
potential receptors are provided in Appendix B.
All source fluid concentrations are assigned a relative concentration of 1 mg/L. As part of the
dose assessment (Section 4), the relative concentrations arriving at the receptor well(s) are
corrected to actual concentrations of the source fluid, and dose calculations are conducted. For
example, if the injection fluid concentration of uranium is 3 mg/L and a relative concentration of
0.01 mg/L reaches a receptor, this would result in a concentration of 0.03 mg/L for the dose
calculation. The breakthrough curves for all of the leak simulations are included in Appendix B.
It should be noted that for all the modeling runs discussed in Chapter 2, no retardation is
considered. Retardation effects are considered in Chapter 4. In the absence of significant
radioactive decay, Retardation only affects the time of the peak dose and not its magnitude.
Catastrophic Leak Failures
As described in NRC 2009b, the Wyoming Department of Environmental Quality (WDEQ)
identified more than 80 spills at the Smith-Ranch Highland site during commercial operations
from 1988 to 2007. This is the largest NRC-licensed ISL uranium recovery facility. The size of
the spills at Smith Ranch-Highland has ranged from a 50- to 100-gallon spill in February 2004 to
a 198,500-gal spill of injection fluid in June 2007. The spills most commonly involved injection
fluids containing 0.5 to 3.0 mg/L uranium, although spills of production fluids containing 10.0 to
152 mg/L uranium also have occurred. These spills have been caused predominantly by the
failure of joints, flanges, and unions of pipelines and at wellheads. The large June 2007 spill at
Smith Ranch-Highland was the apparent result of a failed fitting.
2-17
-------
Nine simulations were conducted (LI through L9) to simulate the potential effects of a
catastrophic failure (Table 2-2). Leak rates for this scenario range from 100,000 to 200,000 and
introduction to the ground water takes place over a 7-day period. Although the largest reported
leak is 200,000 gallons and most of that water was discharged to surface water, it is still
conceivable that it could take 7 days to detect and remedy a similar release. Since the fluid is
released over 7 days, the rate is about 19 gallons per minute (gpm). Other release rates (i.e., 9.5
and 14 gpm) are also investigated. The catastrophic failure simulations also investigate the
effects of hydraulic gradient and conductivities.
2-18
-------
Table 2-2: Catastrophic Leak Failure Scenarios
CATASTROPHIC LEAK FAILURES
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Leak Time
(day)
Model Area (ft2)
Leak Area (ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
328, 656, 1,640,
3,280
53
7
5,280 (ft) x 5,280 (ft)
= lmi2
100
10
10/250
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Leaking Fluid Cone.
(milligram per L)
Hydraulic
Conductivities (ft/day)
Distribution coefficients
(ml/g)
Leakage (gallons)
Time until Peak Arrival
(days)
Relative Peak
Concentration
LI
0.1
20
65
6.5
0.65
1
10
Rf1
Specified in
Spreadsheet
150,000
3.96X101
8.75 xKT3
L2
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
150,000
3.61xl02
6.17X10'3
L3
0.001
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
150,000
4.52xl03
7.18X10'3
L4
0.01
20
65
6.5
0.65
1
1
Rf
Specified in
Spreadsheet
100,000
3.60xl03
4.15X10'3
L5
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
100,000
3.61xl02
4.25 xlO'3
L6
0.01
20
65
6.5
0.65
1
100
Rf
Specified in
Spreadsheet
100,000
3.96X101
4.89xlO'3
L7
0.01
20
65
6.5
0.65
1
1
Rf
Specified in
Spreadsheet
200,000
3.60xl03
7.94xlO'3
L8
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
200,000
3.61xl02
8.01 xlO'3
L9
0.01
20
65
6.5
0.65
1
100
Rf
Specified in
Spreadsheet
200,000
3.96X101
9.48xlO'3
1 Rf = retardation factor calculated in Section 4
2-19
-------
Simulations LI through L3 were focused on evaluating the effects of hydraulic gradients on the
release concentrations and arrival times. The leak rate is 14 gpm (150,000 gallons) and the
hydraulic gradients for LI, L2 and L3 are 0.1, 0.01 and 0.001, respectively. Hydraulic
conductivities for these three runs are all set to the median value investigated (i.e., 10 ft/day). As
shown in Appendix B and presented in Table 2-2, the peak arrival times for simulations LI
through L3 are 40, 361 and 452 days, respectively. Since all parameters are constant except for
hydraulic gradients, the model predictions are as expected and the steeper hydraulic gradients
result in shorter travel times. Furthermore, since the leak volume is creating a recharge mound
and affecting the hydraulic gradients, the arrival times are not linearly scaled (i.e., a factor of 10
increase in gradient does not result in a factor of 10 reduction in travel time). Relative peak
concentrations are 8.75><10~3, 6.17><10~3 and 7.18><10~3. The shape and longevity of the recharge
mound also affect the contaminant concentrations and lead to an apparent discrepancy in the
arrival concentrations in which the maximum relative concentrations observed for a hydraulic
gradient of 0.01 are less than those observed under a gradient of 0.001.
Simulations L4 through L6 were designed to investigate the sensitivity of the results to hydraulic
conductivities (Table 2-2). The assigned values for hydraulic conductivities are 1,10 and
100 ft/day, respectively. A hydraulic gradient of 0.01 ft/ft and a leak rate of 9.5 gpm are common
to all three simulations. The arrival times for simulations L4 through L6 are 3,600, 361 and
39.6 days, respectively (Table 2-2). Since ground water velocities are linearly related to
hydraulic conductivities, there is almost a factor of 10 difference in the peak arrival times. As
was previously observed in simulations LI through L3, the relationship is not exactly linear,
because the recharge causes some mounding and subsequently alters the hydraulic gradients as a
function of the hydraulic conductivities.
Relative peak concentrations for simulations L4 through L6 are 4.15><10~3, 4.25><10~3 and
4.89x 10"3. As indicated by the results, the peak concentrations are slightly higher for the higher
hydraulic conductivities and are related to the greater dispersive effects at the lower hydraulic
conductivities, as observed by the wider breakthrough curves presented in Appendix B.
Simulations L7 through L9 are identical to L4 through L6 except that the leak volume has been
increased to 200,000 gallons (Table 2-2). The arrival times for simulations L7 through L9 are
3,600, 361 and 39.6 days, respectively. The arrival times are identical to those observed for
simulations L4 though L6 and indicate that the additional 100,000 gallons added as recharge do
not appreciably alter the hydraulic gradients. The relative peak concentrations, however, are
about a factor of 2 higher, which is consistent with the increase of twice the mass of contaminant
introduced to the system.
Long-Term/Low-Volume Leak(s)
To evaluate the potential effects of a longer-term leak (3 years) at a lower volume release rate
(i.e., 1 to 2 gpm), nine simulations were conducted (L10 through L18). All of the parameters are
identical to the previous leak simulations except the leak volume and timeframes (Table 2-3).
The 3-year timeframe was selected, because this is the upper range of the times that individual
cells (e.g., in a 5-spot pattern) are mined. It is postulated that underground piping would be either
moved or inspected within this timeframe. A 1- to 2-gallon leak is assumed because this volume
2-20
-------
is sufficiently small to where a long-term leak would possibly go either undetected or
uncorrected.
Simulations L10 through L12 are designed to evaluate the effects of hydraulic gradients on the
release concentrations and arrival times. The leak rate is 1.5 gpm for 3 years, and the hydraulic
gradients forLlO, Lll andL12 are 0.1, 0.01 and 0.001, respectively. Hydraulic conductivities
for these three runs are all set to the median value investigated (i.e., 10 ft/day). The peak arrival
times for simulations L10 through L12 are 353, 1,180 and 4,950 days, respectively (Table 2-3).
As has been previously observed, the steeper hydraulic gradients accelerate the peak arrivals, but
do not behave linearly. As the hydraulic gradients flatten, the leak has a more pronounced impact
on the arrival times. For example, the peak arrival for a gradient of 0.01 ft/ft is approximately 3
times longer than for a peak arrival at a gradient of 0.1 ft/ft, and is about 4 times slower than the
peak arrival at a gradient of 0.001 ft/ft.
Relative peak concentrations for simulations L10 through L12 are 1.11 x 10"2, 6.87x 10"2 and
9.19* 10" , respectively (Table 2-3). As shown in the Appendix B charts, the leading edge of the
breakthrough curve for peak arrivals under a gradient of 0.1 ft/ft is very steep, with the maximum
concentration maintained for about 800 days. The breakthrough curves for runs LI 1 and L12
exhibit more traditional behavior and are depicted by a more gradual arrival, a defined peak, and
gradual tapering off. The difference in this behavior can be explained, because the very steep
hydraulic gradient leads to a compact plume with very small concentration gradients.
Furthermore, the plume that forms under the higher gradients (Figure 2-5) becomes far less
elongated than at lower gradients (Figure 2-6), and the mass is distributed over a larger area,
which results in lower relative peak concentrations.
In runs L13 through L15, the hydraulic gradient is set to 0.01 ft/ft and the leak rate is 1 gpm
(Table 2-3). The hydraulic conductivity, however, is assigned a value of either 1 ft/day (L13),
10 ft/day (LI4) or 100 ft/day (LI5). The peak arrival times are 4,160, 1,180 and 451 days,
respectively. As expected, the lower hydraulic conductivities lead to longer travel times.
The relative peak concentrations for simulations L13 through L15 are 6.02x 10"2, 4.77x 10"2 and
5.92x 10"3, respectively (Table 2-3). The peak concentrations are higher at lower hydraulic
conductivities, because although the same amount of mass has entered the system, the plume
created by the lower hydraulic conductivity remains more compact and is less elongated, as
shown in Figure 2-7 and Figure 2-8.
2-21
-------
Table 2-3: Long-Term/Low Volume Leak Scenarios
LONG-TERM/LOW VOLUME LEAK FAILURES
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Leak Time (yr)
Model Area (ft2)
Leak Area
(pipe - in2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
328, 656, 1,640,
3,280
53
3
5,280 (ft) x 5,280 (ft)
= lmi2
28
10
10/250
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Leaking Fluid Cone.
(milligram per L)
Hydraulic
Conductivities (ft/day)
Distribution coefficients
(ml/g)
Leak Rates
(gallons/min)
Time until Peak Arrival
(days)
Relative Peak
Concentration
L10
0.1
20
65
6.5
0.65
1
10
Rf1
Specified in
Spreadsheet
1.5
3.53xl02
l.llxlQ-2
Lll
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
1.5
l.lSxlO3
6.87xlO'2
L12
0.001
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
1.5
4.95xl03
9.19X10'2
L13
0.01
20
65
6.5
0.65
1
1
Rf
Specified in
Spreadsheet
1
4.16xl03
6.12X10'2
L14
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
1
l.lSxlO3
4.77xlO'2
L15
0.01
20
65
6.5
0.65
1
100
Rf
Specified in
Spreadsheet
1
4.51xl02
5.92xlO'3
L16
0.01
20
65
6.5
0.65
1
1
Rf
Specified in
Spreadsheet
2
4.27xl03
1.04X10'1
L17
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
2
l.lSxlO3
S.SOxKT2
L18
0.01
20
65
6.5
0.65
1
100
Rf
Specified in
Spreadsheet
2
2.95xl02
l.lSxlO'2
Rf = retardation factor calculated in Section 4
2-22
-------
Figure 2-5: Relative Concentrations for L10 at 140 Days
2-23
-------
Figure 2-6: Relative Concentrations for Lll at 140 Days
2-24
-------
,;;::
Figure 2-7: Relative Concentrations for L13 at 877 Days
2-25
-------
Figure 2-8: Relative Concentrations for L15 at 877 Days
2-26
-------
Simulations LI 6 through LI 8 were conducted to investigate the sensitivity of the results to leak
rates (Table 2-3). These runs are paired with simulations L13 through L15 and, with the
exception of leak rate, all of the input parameters for these runs are identical for each pair (i.e.,
L16 and L13, L17 andL14, andL15 and L18). The relative peak arrival time for L16
(4,270 days) is longer than L13 (4,160 days), which may seem counterintuitive since L16 should
have steeper gradients and faster velocities. An inspection of the output files, however, reveals
that the additional volume of water (2 vs. 1 gpm) does increase the velocity, but the increase in
mass redistributes the center of mass leading to a longer time before the peak concentration is
reached. At a hydraulic conductivity of 10 ft/day (LI4 and LI7), the relative peak arrival times
(1,180 days) are virtually the same. The relative peak concentrations, however, vary by a factor
of a little less than 2, with the peak concentration of L14 being 4.77xlO"2, and S.SQxlO"2 for L18.
These relationships indicate that the changes caused to the flow field by the higher injection rates
are balanced by the additional mass. The differences observed between runs LI8 and LI5 are
more pronounced, with the peak arrival time for LI8 being 295 days versus the 451 days for
LI5; and the relative peak concentrations are 1.18xlO~2 and 5.92><10~3 for runs LI8 and LI5,
respectively. These results indicate that, since the plume at higher hydraulic conductivities is
more elongated and there is less movement of the plume in the transverse direction relative to the
longitudinal, the concentration gradients within the plume are less pronounced and lead to a
greater differentiation in travel times and relative concentrations.
Short-Term/High-Volume Leak(s)
The final leak scenario pertains to shorter-term, higher-volume leaks and involves six
simulations (LI9 through L24) (Table 2-4). In these simulations, the release time has been
decreased to 28 days. All of the parameters are identical with the exception of the leak rates,
which vary from 1 gpm to 40 gpm. Although it is very unlikely that 20 and 40 gpm leaks would
continue undetected for 28 days, these leak rates are included to provide an upper bound on the
sensitivity analysis. The hydraulic conductivity is set to 10 ft/day and the hydraulic gradient is
0.01 ft/ft.
The peak arrival times for all six simulations are identical at 382 days, thus indicating that any
changes to the hydraulic gradients are balanced by the increased introduction of mass. The leak
rates for runs L19 and L20 are 1 and 2 gpm, respectively. The relative peak concentration for
L19 is 2.0Qx 10"3, which is approximately one-half of the peak relative concentration for L20 of
3.9Qx 10"3. This nearly linear relationship between the amount of mass released and the relative
peak concentrations is maintained for runs L21 (3 gpm) and L22 (4 gpm) in which the relative
peak concentrations are 5.70x 10"3 and 7.40x 10"3, respectively. For the higher concentrations,
however, the linear relationship is not as apparent and relative peak concentrations for L23
(20 gpm) and L24 (40 gpm) are 3.07x 10"2 and 5.38x 10"2, respectively. All the peak arrival
breakthrough curves are presented in Appendix B.
2-27
-------
Table 2-4: Short-Term/High-Volume Leak Scenarios
SHORT-TERM/ HIGH VOLUME LEAK SCENARIOS
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Leak Time
(days)
Model Area (ft2)
Leak Area
(pipe - in2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
328, 656, 1,640,
3,280
53
28
5,280 (ft) x 5,280 (ft)
= Imi2
28
10
10/250
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Leaking Fluid Cone.
(milligram per L)
Hydraulic
Conductivities (ft/day)
Distribution coefficients
(ml/g)
Leak Rates
(gallons/min)
Time until Peak Arrival
(days)
Relative Peak
Concentration
L19
0.01
20
65
6.5
0.65
1
10
Rf1
Specified in
Spreadsheet
1
3.82xl02
2xlO'3
L20
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
2
3.82xl02
3.9xlO'3
L21
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
3
3.82xl02
5.7xlO'3
L22
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
4
3.82xl02
7.4xlO'3
L23
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
20
3.82xl02
3.07X10'2
L24
0.01
20
65
6.5
0.65
1
10
Rf
Specified in
Spreadsheet
40
3.82xl02
5.38xlO'2
Rf = retardation factor calculated in Section 4
2-28
-------
Summary and Conclusions of Leak Scenarios
A total of three leak and/or spill scenarios were conducted to investigate the potential impacts to
nearby receptors. Nine simulations were performed to evaluate catastrophic leaks. Long-
term/low-volume releases were also simulated with nine simulations, and short-term/high-
volume releases were simulated with six simulations. The sensitivity analysis was focused on
those parameters that are most uncertain. These parameters include hydraulic gradients and
conductivities, leak rates, and durations. All of the concentrations are calculated as relative
concentrations assuming a 1-mg/L source term and will be adjusted to actual concentrations
during the dose assessment (Section 4).
All of the relative peak concentrations versus peak arrival times are shown in Figure 2-9. The
highest relative concentrations are associated with L12, L16 and L17. All three of these
simulations are associated with the long-term/low-volume leaks. Runs L12 and LI6 both have
low ground water velocities. In the case of run L12, this is due to the low gradient (0.001 ft/ft),
whereas the slow ground water velocity for run L16 is because of the low hydraulic conductivity
(1 ft/day). The input for simulation L17 is identical to L16 except that the hydraulic conductivity
in L17 is set to 10 ft/day. This explains the shorter peak arrival time.
Relative Concentrations Versus Time
0.12
0.1
| 0.08
E
4->
OP
g 0.06
u
OJ
(0
0,04
0.02
L16
L17
L12
111
L13
L24
Maximum
Relative
Concentration
L14
L23
0 1000 2000 3000 4000 5000 6000
Time (days)
Figure 2-9: Maximum Relative Concentrations versus Time for All Leak Simulations
2-29
-------
In all cases where the hydraulic conductivity was set to 100 ft/day and/or the hydraulic gradient
was specified as 0.1 ft/ft, the relative concentrations did not exceed 0.012 and the peak arrival
times were less than 452 days. These results are explained by the fact that under more rapid
velocities, the plume becomes more elongated and distributes the mass over a larger volume.
2.6.2 Excursion Scenarios
To investigate the potential effects of excursions beyond the active injection/production wells or
into other non-mined geologic units above and/or below the mined unit, three excursion
scenarios were developed: (1) injection fluid excursions downgradient within the same
lithologic unit, (2) fluid excursions into overlying units through abandoned boreholes, and
(3) fluid excursions into overlying units through discontinuous aquitard(s). The model runs
performed for Scenario 1 are divided into 7 series with each series consisting of 9 simulations for
a total of 63 simulations. Three simulations were conducted to investigate Scenario 2, and two
abandoned borehole simulations were performed (Scenario 3). Summaries of all of the
simulations are provided in Tables 2-5 through 2-11.
Of the 63 simulations, 54 involve either 5- or 7-spot injection\pumping well configurations
(Figure 2-2). The remaining 9 simulate multiple 5-spot well configurations. As discussed in
Section 2.6.2.7, these multiple 5-spot simulations results in longer transport times and lower
peak concentrations.
The results from the single 5- and 7-spot scenarios, however, are more representative of the
mining practices for extracting uranium from isolated outliers and uranium-rich stringers.
The areal extent of the model grid is identical for all of the excursion simulations and consists of
an area that covers 6.25 mi2 (Figure 2-10). Horizontal grid spacing is on 10-ft centers in the
vicinity of the spill, and is gradually increased to a spacing of 1,000 ft at the model boundaries.
The model grid for all of the scenarios is divided into nine layers (Figure 2-11). The vertical grid
spacing, however, is different for Series 1 through 6 of Scenario 1 than it is for Series 7 of
Scenario 1 and for Scenarios 2 and 3. For Series 1 through 6, all the layers are a uniform 75 ft
thick (Figure 2-11). For the 7* series of Scenario 1 and Scenarios 2 and 3, variable layer
thicknesses are assigned and the mined interval thickness is reduced from 75 to 20 ft thick
(Figure 2-11).
2-30
-------
Constant Head Boundaries
Constant Head Boundaries
13300 feet
Figure 2-10: Plan View of the Model Grid for Excursion Scenarios (Series 1 through 7)
2-31
-------
Thickness (feet)
Series 1 -5 Series 6 -7
75 75
Mined Interval
75
Figure 2-11: Cross Section View of the Model Grid for Excursion Scenarios
(Series 1 through 7)
It is assumed that ground water is flowing due south and the model grid is oriented perpendicular
to ground water flow. Constant head boundaries are assigned along the northern and southern
boundaries in order to simulate the regional gradient (Figure 2-10). Boundaries on the east and
west of the model domain are oriented perpendicular to ground water flow (i.e., hydraulic divide)
and, therefore, ground water does not enter or exit the model along these boundaries (i.e., no-
flow boundary conditions). This boundary configuration makes the solutions to the flow problem
more unique than would be obtained if constant head boundaries surrounded the entire domain.
The base of the model is also represented by a no-flow boundary condition.
Since the most common injection/pumping patterns are 5- and 7-spot configurations, both types
of arrangements have been simulated (Figure 2-12 and Figure 2-13, respectively). The extraction
well is always assumed to be in the center with the injection wells along the periphery of the
pattern. Although the injection is only assumed to occur for 3 years, the extraction wells remain
on for the entire simulation and remove between 1% to 3% more than the volume of water being
injected. If the specified pumping rate is 153 gpm, then each of the four injection wells (5-spot
pattern) will be injecting at a rate of 37.5 gpm.
2-32
-------
Injectionl Injections
Si ®
& Pumping
fi ®
Injection 4
JS40 ft
1000 feet
Figure 2-12: Plan View of PumpingMnjection Well Configurations and Receptor
Locations for Excursion Scenarios (Series 1, 2, 5 and 6)
2-33
-------
Injectionl Injection4
Injection2 Si g, & injections
Injection 3 Injection 6
528ft
gi
856ft
1840ft
gl
1000 feet
3480ft
Figure 2-13: Plan View of PumpingMnjection Well Configurations and Receptor
Locations for Excursion Scenarios (Series 3 and 4)
2-34
-------
Potential receptors are assumed to be located at 528 ft, 856 ft, 1,840 ft and 3,480 ft down-
gradient. In the discussion that follows, all of the relative peak arrival times and concentrations
cited are for the potential receptor located at 528 ft down-gradient of the release (Figure 2-14).
Breakthrough curves (i.e., concentrations as a function of time and distance) for all of the
potential receptors are provided in Appendix C. These curves are based on a retardation factor
of 1. Radionuclide-specific retardation effects are considered in Chapter 4.
l.OOE-01
9.00E-02
8.00E-02
7.00E-02
<=
o
2 6.00E-02
<=
o
| 5.00E-02
O
01
'I 4.00E-02
•528 ft
•856ft
1840 ft
•3480ft
3.00E-02
2.00E-02
l.OOE-02
O.OOE+00
O.OOE+00 5.00E+03
1.00E(04 1.50E+04
Time (days)
2.00E+04
2.BOH 104
Figure 2-14: Example of Breakthrough Curves at Receptor Locations
All source fluid concentrations are assigned a relative concentration of 1 mg/L. As part of the
dose assessment (Section 4), the relative concentrations arriving at the receptor well(s) are
corrected to hypothesized concentrations of the source fluid and dose calculations are conducted.
2.6.2.1 Series 1 - 5-Spot 250-ft Spacing
The first series of simulations for Scenario 1 not only investigated the sensitivity of the results to
hydraulic conductivities, gradients and extraction/injection rates, but also allowed an evaluation
of the sensitivity to injection well spacing when compared to Series 2 (Section 2.6.2.2)
counterpart simulations (Table 2-5 and Table 2-6). Since an injection/extraction well spacing of
250 ft is on the upper bound of well spacings typically used, this value was assigned in order to
accentuate the differences in the results when compared against smaller spacings. Of the nine
series of simulations, two series (1 and 3) assumed this upper bound on the well spacing. For the
2-35
-------
remaining series, the wells were spaced at more commonly used intervals between 50 and 150 ft
apart.
The only difference among model runs la, Ib and Ic is that the hydraulic gradients are specified
as either 0.1, 0.01, or 0.001, respectively.
As shown in Appendix C and presented in Table 2-5, the peak arrival times for simulations la
through Ic are 193, 1,080 and 1,600 days, respectively. Since all parameters are constant except
for hydraulic gradients, the model predictions are as expected, and the steeper hydraulic
gradients result in shorter travel times. Furthermore, since the pumping/injection wells are
altering the regional hydraulic gradients, the arrival times are not linearly scaled (i.e., a factor of
10 increase in gradient does not result in a factor of 10 reduction in travel time).
Relative peak concentrations for runs la, Ib and Ic are 6.94xlO~3, 6.54xlO~2 and 9.32xlO~2mg/L
at a receptor well at a distance of 528 ft. The difference in these results can be explained by the
fact that a steeper hydraulic gradient leads to a more elongated plume in which the mass is
distributed over a larger down-gradient area, resulting in lower relative peak concentrations.
Thus, the peak arrival concentration of run la is lower than run Ib, and relative concentrations of
Ib are less than run Ic.
Simulations Id, le and If are designed to investigate the impact that hydraulic conductivity has
on the peak arrival times and relative peak concentrations. Hydraulic conductivities are 1, 10 and
100 ft/day, which coincide with the peak arrival times of 1,700, 1,100 and 193 days,
respectively. As expected, the longer arrival times are associated with the lower hydraulic
conductivities. Also of interest is that the peak arrival times of runs peak arrivals times of Ib and
le are very close (1,080 and 1,100, respectively) and runs Ic and Id are also similar (1,600 and
1,700, respectively). These similarities are related to the fact that velocities will be controlled by
the product of hydraulic conductivity multiplied by the hydraulic gradient. For runs Ib and le,
this product is 1.0, and for runs Ic and Id the product is 0.1.
Relative peak concentrations for runs Id, le and If are 2.14X10"1, 6.55xlO"2 and 6.94xlO"3mg/L.
As has been observed for other simulations, an increase in hydraulic conductivity causes the
plume to become more dispersed and leads to lower relative peak concentrations (Table 2-5).
2-36
-------
Table 2-5: Series 1 - 5-Spot Injection at a Spacing of 250 feet - Receptor Well at 528 feet
SERIES 1 - 5-SPOT INJECTION AT 250 FOOT SPACING
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1,840, 3,480
53
5-Spot (pumping from
center well)
250
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
la
0.1
20
65
6.5
0.65
1
100
Rf1 Specified
in
Spreadsheet
5 -spot -
250
150
153
3
1.93xl02
6.94xlO"3
Ib
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
l.OSxlO3
6.54xlO"2
Ic
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
1.60xl03
9.32xlO"2
Id
0.1
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
l.VOxlO3
2.14X10"1
le
0.1
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
l.lxlO3
6.55xlO"2
If
0.1
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
1.93xl02
6.94xlO"3
lg
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
2.6xl04
1.90xlO"4
Ih
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
1.60xl03
9.27 xlO"2
li
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5 -spot -
250
150
153
3
l.OSxlO3
6.54xlO"2
Rf = retardation factor calculated in Section 4
2-37
-------
Table 2-6: Series 2 - 5-Spot Injection at a Spacing of 50 feet - Receptor Well at 528 feet
SERIES 2 - 5-SPOT INJECTION AT 50 FOOT SPACING
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation Time
(yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area (ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1,840, 3,480
53
5 -Spot (pumping from
center well)
50
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
10/1000
Homogenous
i
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
2a
0.1
20
65
6.5
0.65
1
100
Rf1 Specified
in
Spreadsheet
5-spot - 50
150
153
o
J
1.32xl02
9.62xlO"3
2b
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.05xl03
2.35xlO"2
2c
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.60xl03
8.75 xlO"3
2d
0.1
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
o
J
1.52xl03
LlSxlO"2
2e
0.1
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.05xl03
2.35xlO"2
2f
0.1
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.32xl02
9.62 xlO"3
2g
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
8.22xl03
3.39xlO"6
2h
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.60xl03
8.22xlO"3
2i
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot - 50
150
153
3
1.05xl03
2.35xlO"2
Rf = retardation factor calculated in Section 4
2-38
-------
The hydraulic gradient assigned to simulations Ig, Ih and li is 0.01 ft/ft. The effect that the
regional hydraulic gradient will have on the results is evaluated by comparing simulation output
from these runs to results obtained from runs Id, le and If, which are assigned a regional
gradient of 0.1 ft/ft.
The peak arrival times for simulations Ig, Ih and li are 26,000, 1,600 and 1,080 days,
respectively. The arrival time for run Ig is about 11 times that of run Id (1,700 days), which is
close to what would be expected, since the hydraulic gradient is 10 times lower in run Ig than in
run Id. This same relationship for the peak arrival times for Ih and li, however, are not
maintained when compared to runs le and If, respectively. This is primarily due to the fact that
the injection rates are held constant across all the simulations and the relative effects on
hydraulic gradients will be much greater at lower hydraulic conductivities.
Relative peak concentrations for runs Ig, Ih and li are l.SQxlO"4, 9.27xlO~2 and 6.54xlO~2mg/L.
These results indicate that with flatter gradients (i.e., 0.01), the capture zone for the pumping
wells becomes more pronounced laterally and more of the injectant is captured by the extraction
wells. The inter-relationship between the hydraulic conductivities, regional gradient and
pumping rates is complex, but in general, there is an internal consistency to the results. For
instance, the relative concentrations for runs Ib and le are very similar and, upon inspection of
the data files, it is evident that the product of hydraulic conductivity and gradient is 1.0 for both
simulations. This same relationship is observed for runs le and Ih where the product is equal to
0.1.
2.6.2.2 Series 2 - 5-Spot 50-ft Spacing
The input parameters to all of the Series 2 simulations are identical to those performed in Series
1 with the exception that the spacing between the injection wells is decreased from 250 ft to 50 ft
(Table 2-6).
Relative peak arrival times for runs 2a, 2b and 2c are 132, 1,050 and 1,600 days, respectively.
These values are similar to the peak arrival times for simulations la through le, which are 193,
1,080 and 1,600 days. The fact that run pairs 2b: Ib and 2c: le are so close indicates that the
hydraulic gradients beyond the capture zone are very similar and any excursions will migrate at
very similar rates. The peak arrival times for runs 2a and la, however, show a greater divergence
in which the smaller well spacing and higher hydraulic gradient (0.1) results in a shorter travel
time (i.e., 132 versus 193 days). This is because the larger well spacing affects (flattens) the
hydraulic gradient over a larger area and achieves a better degree of capture, as is confirmed by
comparing the relative peak concentrations.
Relative peak concentrations for runs 2a, 2b and 2c are 9.62xlO"3, 2.35xlO"2and 8.75xlO"3mg/L.
The peak concentration for run 2a is somewhat higher than that observed for run la (i.e.,
6.94x 10"3), indicating that at higher regional gradients, the wider well spacing provides better
capture of the lixiviant. At lower regional hydraulic gradients, however, better capture can be
maintained at smaller well spacings, as evidenced by the lower peak concentrations observed in
runs 2b and 2c when compared to their counterparts.
2-39
-------
Simulations 2d, 2e and 2f are designed to investigate the sensitivity of hydraulic conductivity
(i.e., 1, 10 and 100). Relative peak arrival times for these runs are 1,520, 1,050 and 132 days,
respectively. These results indicate somewhat shorter times than those predicted with the larger
well spacings. Peak arrival times of simulations Id, le and If are 1,700, 1,100 and 193 days,
respectively. These results indicate that the larger well spacing tends to level out the hydraulic
gradients to a greater degree, thus resulting in longer travel times.
Relative peak concentrations for runs 2d, 2e and 2f are l.lSxlO"2, 2.35xlO"2and 9.62xlO"3mg/L,
1 9
and the relative peak concentrations for runs Id, le and If are 2.14*10" , 6.55x 10" and
6.94x 10"3 mg/L. The lower concentrations of 2d and 2e than Id and le indicate that the closer
well spacing leads to better capture and lower excursion concentrations at lower hydraulic
conductivities.
The hydraulic gradient assigned to simulations 2g, 2h and 2i is 0.01 ft/ft and the hydraulic
conductivities are 1, 10 and 100, respectively. Peak arrival times for these simulations are 8,220,
1,600 and 1,050 days. The peak arrival times for counterpart simulations Ig, Ih and li are
26,000, 1,600 and 1,080 days, respectively. Based upon these results, the hydraulic gradients are
most impacted by well spacing at the lower hydraulic conductivities. This is because at higher
hydraulic conductivities, the impacts of injection/withdrawal are in closer proximity to the
pumping/injection well(s). Therefore, a wider well spacing will tend to spread out the effects of
pumping over a larger area, although this does not necessarily mean that the capture of the
lixiviant is greater at a larger well spacing, as is evidenced by the relative peak concentrations.
/- T r\
Relative peak concentrations for runs 2g, 2h and 2i are 3.39x10" , 8.22x10" and 2.35x10" mg/L,
and relative peak concentrations for runs Ig, Ih and li are l.SOxlO"4, 9.27xlO"2 and
6.54xlO"2 mg/L. These results indicate that under moderate hydraulic gradients and at lower
hydraulic conductivities, concentrations of the excursions will be lower. At the higher hydraulic
conductivity (i.e., 100 ft/day), however, and smaller well spacing (run 2i), the excursion reaches
the receptor at higher concentrations as compared to the larger well spacing (run li). Although
the larger well spacing flattens out the gradient and slows the migration more relative to the
smaller well spacing, the capture zone for the smaller well spacing does not extend as far in the
lateral directions (perpendicular to flow) as the larger well spacing. Furthermore, there tends to
be more lixiviant that escapes between the wells at the larger spacing.
2.6.2.3 Series 3 - 7-Spot 250-ft Spacing
The input parameters to all of the Series 3 simulations are identical to those performed in
Series 1 with the exception that the extraction and injection wells are in a 7-spot well
configuration (Figure 2-13).
Relative peak arrival times for runs 3a, 3b and 3c are 235, 1,100 and 1,900 days, respectively
(Table 2-7). These values are very similar to the peak arrival times for simulations la through le,
which are 193, 1,080 and 1,600 days. These results indicate that the 7-spot well configuration
does not significantly alter the travel times as compared to the 5-spot well configuration over a
range of hydraulic gradients (i.e., 0.1, 0.01 and 0.001) and at a hydraulic conductivity of 100
ft/day.
2-40
-------
Table 2-7: Series 3 - 7-Spot Injection at a Spacing of 250 feet - Receptor Well at 528 feet
SERIES 3 - 7-SPOT INJECTION AT 250 FOOT SPACING
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1840, 3480
53
7-spot (pumping from
center well)
250
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
3a
0.1
20
65
6.5
0.65
1
100
Rf1 Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
2.35xl02
2.78xlO"3
3b
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
l.lxlO3
3.53xlO'2
3c
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
1.90xl03
6.79xlO"2
3d
0.1
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
3
1.97xl03
1.41X10'1
3e
0.1
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
l.lxlO3
3.53xlO'2
3f
0.1
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
2.35xl02
2.78xlO"3
3g
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
o
J
4.00xl04
1.06xlO"4
3h
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
3
1.90xl03
6.75 xlO"2
3i
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 250
150
153
3
LlOxlO3
3.53xlO'2
Rf = retardation factor calculated in Section 4
2-41
-------
^9 9
Relative peak concentrations for runs 3a, 3b and 3c are 2.78x10" ,3.53x10" and 6.79x10" mg/L,
and relative peak concentrations for runs la, Ib and Ic are 6.94x 10"3, 6.54x 10"2 and
9.32xlO"2 mg/L. Within each series, the flatter regional hydraulic gradients result in higher
concentrations due to the plumes being less elongated and the mass more confined. The 7-spot
well configuration, however, results in lower relative peak concentrations for all of the runs. This
is because the additional pumping/extraction wells allow more overlap of the capture zone(s). In
a 7-spot pattern, the distance from an injection to an extraction well is 250 ft, while for a 5-spot
pattern it is 176 ft.
Hydraulic conductivities for runs 3d, 3e and 3f are 1, 10 and 100 ft/day, which coincide with
relative peak arrival times of 1,970, 1,100 and 235 days, respectively. Relative peak arrivals for
simulations Id, le and If are 1,700, 1,100 and 193 days. Therefore, the peak arrival times for
both sets of simulations are relatively similar over a range of hydraulic conductivities.
Relative peak concentrations for runs 3d, 3eand3fare 1.41X10"1, 3.53xlO"2and2.78xlO"3mg/L,
and relative peak concentrations for runs Id, le and If are 2.14X10"1, 6.55xlO"2 and
6.94xlO"3 mg/L. Lower concentrations are correlated to the higher hydraulic conductivities. As
was previously observed for runs 3a-c and la-c, the 7-spot configuration results in lower relative
peak concentrations for all of the simulations.
The hydraulic gradient assigned to simulations 3g, 3h and 3i is 0.01 ft/ft and the hydraulic
conductivities are 1, 10 and 100, respectively. Peak arrival times for these simulations are
40,000, 1,900 and 1,100 days. The peak arrival times for counterpart simulations Ig, Ih and li
are 26,000, 1,600 and 1,080 days, respectively. Based upon these results, the travel times at the
higher hydraulic conductivities are very similar for the hydraulic conductivity of 1 ft/day (run 3g
and Ig), however, the time until peak arrival for the 7-spot well configuration is considerably
longer. This observation is primarily due to the effect that the additional wells have on flattening
out the hydraulic gradient.
Relative peak concentrations for runs 3g, 3h and 3i are 1.06xlO"4, 6.75xlO"2and 3.53xlO"2 mg/L,
and relative peak concentrations for runs Ig, Ih and li are l.SQxlO"4, 9.27xlO"2 and
6.54x 10"2 mg/L. These results indicate that under moderate hydraulic gradients and at lower
hydraulic conductivities, concentrations of the excursions will be lower. Furthermore, the
relative peak concentrations for all of the 7-spot well configurations are lower than those for the
analogous 5-spot well simulations.
2.6.2.4 Series 4 - 7-Spot 50-ft Spacing
The input parameters to all of the Series 4 simulations are identical to those performed in
Series 3 with the exception that the extraction and injection wells are spaced at 50 ft instead of at
250 ft.
Relative peak arrival times for runs 4a, 4b and 4c are 186, 1,100 and 1,450 days, respectively
(Table 2-8). For comparable simulations in Series 3, the relative peak arrival times for runs 3a,
3b and 3c are 235, 1,100 and 1,900 days, respectively (Table 2-7). These results indicate that at
the high (4c:3c) and low (4a:3a) hydraulic gradients, the relative peak arrival times are shorter
2-42
-------
for the more narrowly spaced wells. This relationship was also observed when the output for
Series 1 and 2 was compared (Section 2.6.2.2) and is caused by greater flattening of the
hydraulic gradient over a larger area with the wider well spacing. At the moderate gradient
(0.01 ft/ft), however, the peak arrival times for runs 3b and 4b are essentially the same. This is
because at the moderate gradient, there is a balance between the regional gradient and the
gradients caused by the pumping/injection in which the net gradients that result are not as
sensitive to the well spacing.
Relative peak concentrations for runs 4a, 4b and 4c are 1.25xlO"2, 4.03xlO"3and 9.01xlO"3 mg/L,
and relative peak concentrations for runs 3a, 3b and 3c are 2.78xlO"3, 3.53xlO"2and
6.79x 10"2 mg/L. The peak concentration for run 4a is somewhat higher than that observed for run
3a, indicating that at higher regional gradients, the wider well spacing provides better capture of
the lixiviant. At lower regional hydraulic gradients, however, better capture can be maintained at
smaller well spacings, as evidenced by the lower peak concentrations observed in runs 4b and 4c
when compared to their counterparts. This same relationship was observed when Series la, b,
and c was compared to Series 2a, b, and c (Section 2.6.2.2).
Hydraulic conductivities for runs 4d, 4e and 4f are 1, 10 and 100 ft/day, which coincide with
relative peak arrival times for runs of 1,390, 1,020 and 186 days, respectively. Relative peak
arrival times for runs 3d, 3e and 3f are 1,970, 1,100 and 235 days. As is expected, the longer
arrival times are associated with the lower hydraulic conductivities. The peak arrival times of 4e
and 3e are very close (1,020 and 1,100, respectively). For the remaining simulations, however,
the results are similar, but do not compare as well. Due to the hydraulic gradients not being
affected over as large an area, the smaller well spacing results in shorter peak arrival times.
Relative peak concentrations for runs 4d, 4e and 4f are 3.07xlO"2, 4.03xlO"2 and 1.25xlO"2 mg/L,
and relative peak concentrations for runs 3d, 3eand3fare 1.41X10"1, 3.53xlO"2and
2.78xlO"3 mg/L. With exception of the 4e:3e run comparison, the closer well spacing leads to
lower relative peak concentrations. The closer well spacing also leads to better capture, in
general, as evidenced by the similar release concentrations among runs 4d, 4e and 4f
The hydraulic gradient for runs 4g, 4h, and 4i is 0.01 ft/ft and the hydraulic conductivities are 1,
10 and 100 ft/day, respectively. Relative peak arrival times for runs 4g, 4h and 4i are 7,840,
1,450 and 1,100 days, respectively (Table 2-8). For comparable simulations in Series 3, the
relative peak arrival times for runs 3g, 3h and 3i are 40,000, 1,900 and 1,100 days. At the lower
hydraulic conductivities (4g and 4h), the closer well spacing results in shorter arrival times. For a
hydraulic conductivity of 100 ft/day, the arrival times (4i:3i) are identical.
2-43
-------
Table 2-8: Series 4 - 7-Spot Injection at a Spacing of 50 feet - Receptor Well at 528 feet
SERIES 4 - 7-SPOT INJECTION AT 50 FOOT SPACING
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1,840, 3,480
53
7-Spot (pumping from
center well)
50
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
4a
0.1
20
65
6.5
0.65
1
100
Rf1 Specified
in
Spreadsheet
7-spot - 50
150
153
o
J
1.86xl02
1.25xlO"2
4b
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
LlOxlO3
4.03 xKT3
4c
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
1.45xl03
9.01 xlO'3
4(1
0.1
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
o
J
1.39xl03
3.07xlO"2
4e
0.1
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
1.02xl03
4.03 xlO'2
4f
0.1
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
1.86xl02
1.25 xlO"2
4g
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
7.84xl03
2.14xlO"6
4h
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
1.45xl03
8.94xlO"3
4i
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
7-spot - 50
150
153
3
LlOxlO3
4.03 xlO'2
Rf = retardation factor calculated in Section 4
2-44
-------
Relative peak concentrations for runs 4g, 4h and 4i are 2.14* 10"6, 8.94x 10"3 and 4.03 x 10"2 mg/L.
These outcomes indicate that capture of the injectant is more effective at the lower hydraulic
conductivities and results in lower relative peak concentrations. A comparison against
simulations 4d, 4e and 4f indicates that the steeper hydraulic gradient results in larger releases
because the lateral capture is not as effective.
Relative peak concentrations for runs 4g, 4h and 4i are lower than those observed for
4 9 9
runs 3g, 3h and 3i, which are 1.06x10" , 6.75x10" and 3.53x10" mg/L. These results support
earlier findings indicating that the closer well spacing captures more of the injectant.
2.6.2.5 Series 5 - 5-Spot Injection/Pumping Rates Dependent Upon Hydraulic Conductivity
The injection/pumping rates specified in all of the simulations performed in Series 1 through 4
are 150 gpm (injection) and 153 gpm (pumping). Although these rates are the same, in actuality,
the hydraulic conductivities would be considered in determining the amount of water that is
pumped and injected. This correlation was not factored into the Series 1 through 4 simulations
because it would make it very difficult to isolate the effects of the other parameters (e.g.,
hydraulic gradients, well configuration and spacing). For Series 5, however, the
injection/pumping rates have been adjusted to more realistically reflect the hydraulic
conductivity of the system.
To estimate pumping/injection rates at hydraulic conductivities of 1, 10 and 100 ft/day, a
constant head boundary was set to an elevation representative of the approximate pumping level
in each of the wells. MODFLOW output files provide the amount of water removed by the
constant head boundary, and this value was subsequently used as input for the pumping/injection
rates. The specified rates for hydraulic conductivities of 1, 10 and 100 ft/day are 7.15/7, 51/50
and 510/500 gpm, respectively (Table 2-9).
In addition to the change in the pumping/extraction rates, the well spacing for the Series 5
simulations is set to 150 ft, as opposed to the 50 or 250 spacings assigned in the earlier
simulations.
2-45
-------
Table 2-9: Series 5 - 5-Spot Injection\Pumping Rates Dependent Upon Hydraulic Conductivity
SERIES 5 - 5-SPOT INJECTION/PUMPING RATES DEPENDENT UPON HYDRAULIC CONDUCTIVITY
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1840, 3480
53
5-Spot (pumping from
center well)
150
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
5a
0.1
20
65
6.5
0.65
1
10
Rf1 Specified
in
Spreadsheet
5-spot- 150
50
51
o
J
LlOxlO3
3.76xlO"2
5b
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.52xl03
6.79xlO"2
5c
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
l.SxlO4
8.24 xlO"4
5d
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot- 150
7
7.15
o
J
l.OlxlO4
1.53xlO'2
5e
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.52xl03
6.79xlO"2
5f
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
3
1.05xl03
l.OlxKT1
5g
0.001
20
65
6.5
0.65
1
1
Rf Specified
in Spreadsheet
5-spot- 150
7
7.15
o
J
1.28xl04
l.OSxlO"16
5h
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
o
J
l.SxlO4
8.24xlO"4
5i
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
o
J
LlOxlO3
l.OSxlO"1
Rf = retardation factor calculated in Section 4
2-46
-------
The hydraulic gradients for runs 5a, 5b and 5c are 0.1, 0.01 and 0.001 ft/ft. All of the hydraulic
conductivities are set to 10 ft/day with an accompanying pumping/injection rate of 51/50 gpm.
Relative peak arrival times for runs 5a, 5b and 5c are 1,100, 1,520 and 18,000 days, respectively
(Table 2-9). Since so many parameters have been changed from those assigned to simulations in
the previous series, it is difficult to draw inter-series comparisons. The arrival time of
18,000 days, however, is one of the longest times observed.
Relative peak concentrations for runs 5a, 5b and 5c are 3.76xlO"2, 6.79xlO"2and 8.24xlO"4 mg/L
and indicate that the capture zone is most effective at the lowest hydraulic gradient of 0.001 ft/ft
(run 5c). Results for runs 5a and 5b indicate that although the capture is more effective at a
gradient of 0.01 than 0.1 ft/ft, the injectant becomes more widely dispersed in the down-gradient
direction under the 0.1 ft/ft gradient due to the higher velocities.
Runs 5d, 5e and 5f are all performed with a hydraulic gradient of 0.01 ft/ft; hydraulic
conductivities of 1, 10 and 100 ft/day; and with the pumping/injection rates varying as a function
of the hydraulic conductivity. Relative peak arrival times for runs 5d, 5e and 5f are 10,100, 1,520
and 1,050 days. The arrival time for 5d is almost 10 times that of 5a, which is expected since the
hydraulic gradient is decreased by a factor of 10. This result also reflects the fact that the
pumping rate for 5d of 7.15 gpm was similarly scaled to a hydraulic conductivity of 1, as was the
pumping rate of 50 gpm to a hydraulic conductivity of 10 ft/day.
Relative peak concentrations for runs 5d, 5e and 5f are 1.53xlO"2, 6.79xlO"2and l.OlxlO"1 mg/L.
The concentrations increase as a function of increasing pumping/injection rates and hydraulic
conductivities and indicate that capture is more complete at the lower hydraulic conductivities.
These results also suggest that the bleed rate (i.e., difference between pumping and injection
rates) should be increased as a function of hydraulic conductivity in order to increase the capture
zone at hydraulic conductivites.
The input for runs 5g, 5h and 5i are identical to that for runs 5d, 5e and 5f, except that the
hydraulic gradient is set to 0.001 ft/ft instead of 0.01 ft/ft. The peak arrival times for runs 5g, 5h
and 5i are 12,800, 18,000 and 1,100 days. With the exception of runs 5h and 5e, the arrival times
for the 5g-e and 5i-f pairs are very similar. This suggests that the hydraulic gradients created by
the pumping/injection wells are large enough to overwhelm the differences in the regional
hydraulic gradients. With respect to run 5h, it appears that the long peak arrival time observed is
due to the very low relative peak concentration (i.e., l.OSxlO"16) in which numerical dispersion
may be a significant contributor.
Relative peak concentrations for runs 5g, 5h and 5i are l.OSxlO"16, 8.24xlO"4and l.OSxlO"1 mg/L.
These results, when compared against those obtained for runs 5d, 5e and 5f, indicate that the
flatter gradients allow better capture and that at the conductivity of 100 ft/day, the injection/
pumping is large enough to overwhelm the differences in regional gradients.
2-47
-------
2.6.2.6 Series 6 - 5-Spot 20-ft Thick Mined Interval
All of the simulations performed in Series 6 are identical to those conducted for Series 5 except
that the model layers vary in thickness and the mined interval was reduced from a thickness of
70 ft to 20 ft (Figure 2-11).
The relative peak arrival times for run 6a, 6b and 6c are 1,700, 1,100 and 20,600 days,
respectively (Table 2-10). As described in Section 2.6.2.5, the relative peak arrival times for runs
5a, 5b and 5c are 1,100, 1,520 and 18,000 days. The effect of pumping/injection on hydraulic
gradients is strongly affected by the transmissivity (i.e., hydraulic conductivity multiplied by
thickness) of the geologic units. The lower transmissivity results in shorter times to peak arrivals
at the low and high gradients (runs 6a and 6c) and a longer time at the medium gradient (run 6b).
These relationships are all related to how the regional and localized gradients created by the
pumping/injection interact to form a capture zone. It also illustrates the complexity and need to
understand the geology and flow system, since the effects of the interactions are not always
intuitive.
Relative peak concentrations for runs 6a, 6b and 6c are S.SlxlO'1, 4.75X10'1 and 1.30><10~3 mg/L
and, as was observed with runs 5a, 5b and 5c, indicate that the capture zone is most effective at
the lowest hydraulic gradient (0.001 ft/ft). All of the relative concentrations are approximately an
order of magnitude higher than the relative peak concentrations for runs 5a, 5b and 5c (i.e.,
3.76x 10"2, 6.79x 10"2 and 8.24x 10"4 mg/L). The higher concentration at the lower transmissivity
is due to the injectant being concentrated within a smaller volume.
Runs 6d, 6e and 6f are all performed with a hydraulic gradient of 0.01 ft/ft; hydraulic
conductivities of 1, 10 and 100 ft/day; and with the pumping/injection rates varying as a function
of the hydraulic conductivity. Relative peak arrival times for runs 6d, 6e and 6f are 9,930, 1,520
and 1,050 days. The times are very similar for runs 6d-5d and essentially identical for runs 6e-5e
and 6f-5f. Therefore, for the same regional gradient, the peak arrival times are insensitive to the
transmissivity.
Relative peak concentrations for runs 6d, 6e and 6f are 1.06X10"1, 4.75X10"1 and 6.45X10"1 mg/L,
and, as was observed with the 6a, b, and c versus 5a, b, and c series, are all considerably higher
than their counterparts where the relative peak concentrations for runs 5d, 5e and 5f are
1.53xlO"2, 6.79xlO"2 and l.OlxlO^mg/L. These results demonstrate that the lower
transmissivities will result in higher relative concentrations.
The peak arrival times for runs 6g, 6h and 6i are 288,000, 20,700 and 1,210 days. The very long
peak arrival time for run 6g and low peak concentration (i.e., 7.4Qx 10"5) indicate that capture at
the lowest hydraulic conductivity is essentially complete. The relative peak arrival times for 5h
and 5i (i.e., 18,000 and 1,100 days) are very close to those observed for runs 6h and 6i, which
indicate that similar hydraulic gradients are being created for both sets of simulations.
2-48
-------
Table 2-10: Series 6 - 5-Spot 20-Foot Thick Mined Interval
SERIES 6 - 5-SPOT - 20-FOOT THICK MINED INTERVAL
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528 856 1 840 3 480
53
5-Spot (pumping from
center well)
150
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
The thickness of layer 5 is
decreased to 20 ft thick and
has the monitoring/
injection/pumping wells.
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
6a
0.1
20
65
6.5
0.65
1
10
Rf1 Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.70xl03
3.3 IxKT1
61)
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
LlOxlO3
4.75X10'1
6c
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
2.06xl04
1.30xlO'3
6(1
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot- 150
7
7.15
3
9.93xl03
1.06X10'1
6e
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.52xl03
4.75X10'1
6f
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
3
1.05xl03
6.45X10'1
6g
0.001
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot- 150
7
7.15
3
2.88xl05
7.40 xlO'5
6h
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
2.07xl04
l.SOxlO-3
6i
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
3
1.21xl03
6.26X10'1
Rf = retardation factor calculated in Section 4
2-49
-------
Relative peak concentrations for runs 6g, 6h and 6i are 7.40xlO~5, l.SQxlO"3 and 6.26X10"1 mg/L.
These results, when compared against those obtained for runs 5g, 5h and 5i (i.e., l.OSxlO"16,
8.24xlO~4and l.OSxlO"1), are all higher and, as was observed between the 6d, e, and f and 5d, e,
and f comparisons, the lower transmissivities result in higher relative concentrations.
2.6.2.7 Series 7 - Twenty-Five 5-Spot Pumping/Injection Cells
To evaluate the potential cumulative effects of multiple pumping/injection cells, the simulations
in Series 7 consist of twenty-five 5-spot injection cells (Figure 2-15). With the exception of the
multiple cells, all of the other input parameters for Series 7 are identical to those for Series 6,
where only one 5-spot pattern was modeled (Table 2-11).
The only difference among model runs 7a, 7b and 7c is that the hydraulic gradients are specified
as 0.1, 0.01 or 0.001, respectively. The relative peak arrival times for runs 7a, 7b, and 7c are
1,090, 4,830 and 1,280 days, while the relative peak arrival times for runs 6a, 6b and 6c are
1,700, 1,100 and 20,600 days. The observed differences are associated with the degree to which
the multiple cells affect the hydraulic gradients as compared to the single pumping/injection cell.
At the highest gradient (runs 7a and 6a), both simulations impact the hydraulic gradients to a
relatively similar degree. At the moderate gradient, however, the multiple cells are more
effective at flattening the hydraulic gradients, which is reflected in the longer peak arrival times
in run 7b as compared to run 6b. For the runs with the relatively flat regional gradients (7c and
6c), the impact of the pumping/injection for the multiple cells in run 7c overwhelms the effects
of the regional gradient, as expressed by the very long travel time in run 6c.
Relative peak concentrations for runs 7a, 7b and 7c are 3.12X10"1, 6.61xlO"4 and 2.7QxlO"4 mg/L,
and the relative peak concentrations for runs 6a, 6b and 6c are 3.31 x 10"1, 4.75x 10"1 and
1.30x 10"3 mg/L. Except for the simulations with the steep regional gradient (7a and 6a), the peak
relative concentrations with the multiple cells are orders of magnitude lower than the single cell
simulations. These results indicate that at lower hydraulic gradients, the majority of the injectant
that is not recaptured is lost from the injection wells that are located to the furthest east and west
and not very much is lost from the line of injection wells located in between. Under the steeper
hydraulic gradient, however, the capture is not complete along the entire line of wells running
east-west and, therefore, the relative peak concentrations for runs 7a and 6a are at similarly
elevated levels.
Runs 7d, 7e and 7f are all performed with a hydraulic gradient of 0.01 ft/ft, and hydraulic
conductivities of 1, 10 and 100 ft/day. Relative peak arrival times for runs 7d, 7e and 7f are
48,700, 4,800 and 1,100 days. At the lower hydraulic conductivities, these results indicate slower
velocities than the comparable runs in Series 6 (6d and 6e), which are 9,930, and 1,520 days and
indicate the formation of flatter gradients formed by the array of pumping/injection wells. At the
hydraulic conductivity of 100 ft/day, however, the arrival time for run 7f is nearly the same as
that for run 6f (1,050 days), indicating that the gradients are less affected by the
pumping/injection at higher hydraulic conductivities.
2-50
-------
a ® 0 0
§ @ ® 8 g
3 ® @ 8
a s ^IniedionB Injection
® @ 8 @ _ @ .
Purnpmg
@ ® ® 0 8
Injection Injection
500 feet
528ft
856ft
1840ft
3480 ft
Figure 2-15: Plan View of the Twenty-five 5-spot Pumping/Injection Well Configurations
and Receptor Locations for Excursion Scenario (Series 7)
2-51
-------
Table 2-11: Series 7 - Twenty-five 5-Spot Pumping/Injection Cells
SERIES 7 - TWENTY-FIVE (25) 5-SPOT PUMPING/INJECTION CELLS
General Assumptions - All Runs
Receptor
Locations (ft)
Simulation
Time (yr)
Well
Configuration
Well Spacing
(ft)
Injection Time
(yr)
Model Area
(ft2)
Model Layers
Minimum/
Maximum Grid
Spacing (ft)
All Layers
528, 856, 1,840, 3,480
53
5-Spot (pumping from
center well)
150
3
13,300 (ft) x 13,300 (ft)
= 6.25 mi2
9
The thickness of layer 5 is
decreased to 20 ft thick and
has the monitoring/
injection/pumping wells.
10/1000
Homogenous
Run
Regional Hydraulic
Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Hydraulic Conductivities
(ft/day)
Distribution coefficients
(ml/g)
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL
Well Pattern (yrs)
Time until Peak Arrival
(days)
Relative Peak
Concentration
7a
0.1
20
65
6.5
0.65
1
10
Rf1 Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.09xl03
3.12X10'1
7b
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
4.83xl03
6.61 xlO'4
7c
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.28xl03
2.70 xlO'4
7d
0.01
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot- 150
7
7.15
3
4.87xl04
2.35xlO'5
7e
0.01
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
4.80xl03
6.61 xlO'4
7f
0.01
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
3
LlOxlO3
LlOxlO'2
7g
0.001
20
65
6.5
0.65
1
1
Rf Specified
in
Spreadsheet
5-spot- 150
7
7.15
3
6.63xl03
5.25 xlO'5
7h
0.001
20
65
6.5
0.65
1
10
Rf Specified
in
Spreadsheet
5-spot- 150
50
51
3
1.28xl03
2.70xlO'4
7i
0.001
20
65
6.5
0.65
1
100
Rf Specified
in
Spreadsheet
5-spot- 150
500
510
3
LlOxlO3
4.50xlO'4
Rf = retardation factor calculated in Section 4
2-52
-------
Relative peak concentrations for runs 7d, 7e and 7f are 2.35xlO~5, 6.61xlO~4and l.lQxlO"2 mg/L
and are all significantly lower than for their Series 6 counterparts, where relative peak
concentrations for runs 6d, 6e and 6f are 1.06X10"1, 4.75X10'1 and 6.45xlO~1mg/L. These results
indicate that, not only is the degree of capture a function of the hydraulic conductivity, but also
that a significant amount of the injectant that is not captured is from the injection into the wells
located farthest to the east and to the west. This conclusion is based on the fact that the receptor
well(s) are placed on the plume centerline and, therefore, in the case with the single injection
well, the concentration will be impacted by any uncaptured releases from the wells to the east,
west and between the pumping and injection wells. For the multiple injection array, however, the
receptor well is farther from the injection wells to the east and west, and therefore, the injectant
concentrations detected at the well are not as high.
The model input parameters for runs 7g, 7h and 7i are identical to those assigned in runs 7d, 7e
and 7f except that the hydraulic gradient was changed from 0.01 to 0.001 ft/ft. The peak arrival
times for runs 7g, 7h and 7i are 6,630, 1,280 and 1,100 days and, except at the highest hydraulic
conductivity, are much shorter than for runs 6g, 6h and 6i, which are 288,000, 20,700 and
1,210 days. These results indicate that at the low and moderate hydraulic conductivities (1 and
10 ft/day) and flat gradients (0.001), the multiple well configuration does not flatten the
hydraulic gradients as much as the single pumping/injection cell configuration.
Relative peak concentrations for runs 7g, 7h and 7i are 5.25xlO"5, 2.7QxlO"4and 4.5QxlO"4 mg/L.
These values are all lower than the relative peak concentrations for runs 6g, 6h and 6i, which are
7.4QxlO"5, l.SQxlO"3 and 6.26X10"1 mg/L. These results are explained by the larger capture zone
that is created by the array of pumping/extraction wells.
While simulations with a single 5- or 7-spot pattern are useful in understanding the interactions
and sensitivity to various modeling parameters, this modeling approach produces conservative
results (i.e., high relative concentration at the receptor well). In general, modeling of multiple
injection/extraction patterns, which are a closer approximation to a full-scale wellfield, results in
lower relative concentrations.
2.6.2.8 Abandoned Borehole Pathway
Several modeling simulations were performed to investigate the potential impacts on overlying
aquifers of an exploratory borehole that penetrates into the mined unit (Table 2-12). The major
conceptual components of the modeled system are a 60-ft thick low-conductivity confining unit
(lx 10"6 ft/day) that separates a 30-ft thick mined interval from an overlying aquifer with an
assigned thickness of either 50 or 100 ft, depending upon the simulation. The abandoned
borehole hydraulically connects the overlying aquifer to the mined unit through a 1-ft2 high
hydraulic conductivity damaged rock zone created as the borehole was cored.
To simulate this system, a 5-layer model was constructed that covers approximately a 1-mi2 area
(Figure 2-16). Grid spacing in the horizontal direction ranges from 1 to 50 ft and in the vertical
direction from 20 to 100 ft (Figure 2-17). The aquifer and mined unit are represented by single
layers and are separated by an aquitard that is divided into 3 layers.
2-53
-------
Table 2-12: Abandoned Borehole Simulations
ABANDONED BOREHOLE SIMULATIONS
General Assumptions - All Runs
Receptor Locations (ft)
Simulation Time (yr)
Well Configuration
Well Spacing (ft)
Injection Time (yr)
Model Area (ft2)
Model Layers
Minimum/ Maximum
Grid Spacing (ft)
All Layers
Disturbed Rock Zone
Around Borehole (ft2)
528, 856, 1,840
53
5 -Spot (pumping from
center well)
150
3
5,280 (ft) x 5,280 (ft)
= 1 mi2
5
The thickness of the mined
interval (layer 5) is 30 ft and
has the monitoring/
injection/pumping wells.
1/50
Confining units
separate upper aquifer
from production zone
1
Run
Regional Hydraulic Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid Cone.
(milligram per L)
Upper Aquifer 1 Hydraulic
Conductivity (ft/day)
Upper Aquifer Thickness (ft)
Mined Interval Hydraulic
Conductivity (ft/day)
Retardation Factor
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL Well
Pattern (yrs)
Time until Peak Arrival (days)
Relative Peak Concentration
AB-R1
0.01
20
65
6.5
0.65
1
100
100
10
1
5-spot - 100
50
51
o
J
2.98xl03
5.81XKT1
AB-R2
0.01
20
65
6.5
0.65
1
10
100
10
1
5-spot - 100
50
51
o
J
4.15xl03
5.12X10"1
AB-R3
0.01
20
65
6.5
0.65
1
100
50
10
1
5-spot - 100
50
51
3
2.06xl03
S.OOxlO"1
2-54
-------
Constant Head Boundaries
Constant Head Boundaries
Receptor Wells
5200 feet
Figure 2-16: Plan View of the Model Grid for the Abandoned Borehole and Discontinuous
Confining Bed Excursion Simulations
2-55
-------
TWckness (Feet|
Figure 2-17: Cross-Sectional View of the Model Grid for the Abandoned Borehole and
Discontinuous Confining Bed Excursion Simulations
Three model simulations were performed to investigate the effects of aquifer thickness and
aquifer conductivity. It is recognized that the results would be influenced by many parameters
(e.g., size and permeability of the damaged rock zone), but the primary objective was to assess
whether, under a reasonable set of assumptions, vertical migration via a borehole could result in
significant releases. It should be kept in mind, however, that for this type of release to occur,
there has to be an upward vertical gradient and the damaged rock zone has to be of sufficient
hydraulic conductivity to allow vertical movement.
The peak arrival time for the base case abandoned borehole simulation (AB-R1) is 2,980 days
and the peak concentration is S.SlxlO'1. This is one of the most significant releases predicted
across all of the potential failure scenarios (i.e., leaks and excursions).
The effect that a lower aquifer hydraulic conductivity has on the results is evaluated in run
AB-R2 where the hydraulic conductivity is lowered from 100 to 10 ft/day. The peak arrival time
for this simulation (AB-R2) is 4,150 days and the peak concentration is 5.12* 10"1. The travel
time has increased from the base case because of the lower ground water velocities. The peak
2-56
-------
concentrations, however, remain very similar. In previous excursion simulations, the higher
hydraulic conductivities typically had lower peak concentrations. In this comparison, however,
the higher hydraulic conductivity has a higher release. This relationship occurs because the
vertical gradient is affected by the hydraulic conductivity of the overlying aquifer. At higher
hydraulic conductivities, the vertical gradient allows more injectant to enter the aquifer via the
borehole. The amount of water entering the upper aquifer also impacts the horizontal gradients
within the aquifer.
The impact of the reduction in aquifer thickness from 100 to 50 ft is investigated in run AB-R3.
The peak arrival time for this simulation is 2,060 days and the peak concentration is S.OOx 10"1.
The shorter peak arrival time can be explained by the steeper hydraulic gradients in the aquifer,
which are due to the water flowing up the borehole encountering a lower transmissivity
(hydraulic conductivity multiplied by thickness). The higher relative concentration occurs
because the amount of injectant entering the aquifer is distributed over a smaller vertical area (or
volume). Breakthrough curves are included in Appendix D.
2.6.2.9 Confining Bed Discontinuity
To investigate the potential effects of a discontinuous confining unit between the mined aquifer
and the overlying aquifer, two simulations were performed (Table). For both of these
simulations, the model grid, domain and input parameters are identical to those of the abandoned
borehole base case (AB-R1), except that the abandoned borehole is replaced by an area of either
10 or 100 ft2, where the hydraulic conductivity is high (100 ft/day). Conceptually, this area is
representative of an erosional or deposit!onal surface that has been filled with the material from
the overlying aquifer (e.g., a sand lens).
For run CBD-R1, this area of discontinuity is specified as 100 ft2. The peak arrival time for this
simulation is 11,900 days and the peak concentration is 5.09x 10"1. This longer arrival time from
that observed for the abandoned borehole simulations occurs because the high transmissivity
zone (i.e., damaged rock zone) does not extend through the entire aquifer. Therefore, injectant
entering the aquifer must do so only at the base, and it takes longer for the injectant to mix with
the water in the aquifer to reach a peak concentration. The peak concentration, however, is very
similar to those observed in runs AB-R1 and AB-R2.
The area of discontinuity was decreased from 100 ft2 to 25 ft2 in run CBD-R2. The peak arrival
time for this simulation is 45,700 days and the peak concentration is 4.73 x 10"1. The longer peak
arrival time for the smaller area is related to the smaller amount of injectant that is migrating
vertically upward. The effect of the smaller area is also reflected in the lower relative peak
concentration. Breakthrough curves are included in Appendix D. Since the potential releases
through the discontinuous layers result in uranium concentrations that are almost the same as
those observed in the wellfield, there was no need to calculate the health-based standards for
these scenarios.
Furthermore, although this analysis was focused upon overlying aquifers, similar results would
be obtained from potential excursions to underlying aquifers.
2-57
-------
Table 2-13: Confining Bed Discontinuity Simulations
CONFINING BED DISCONTINUITY SIMULATIONS
General Assumptions- All Runs
Receptor Locations (ft)
Simulation Time (yr)
Well Configuration
Well Spacing (ft)
Injection Time (yr)
Model Area (ft2)
Model Layers
Minimum/ Maximum
Grid Spacing (ft)
All Layers
528, 856, 1,840
53
5 -Spot (pumping
from center well)
150
3
5,280 (ft) x 5,280 (ft)
= 1 mi2
5
The thickness of the mined
interval (layer 5) is 30 ft
and has the monitoring/
injection/pumping wells.
1/50
Confining units
separate upper aquifer
from production zone
Run
Regional Hydraulic Gradient
Effective Porosity (%)
Dispersivity (ft)
Longitudinal
Transverse
Vertical
Injection Fluid
Cone, (milligram per L)
Upper Aquifer 1 Hydraulic
Conductivity (ft/day)
Upper Aquifer Thickness (ft)
Mined Interval Hydraulic
Conductivity (ft/day)
Area of Confining Bed
Discontinuity (ft2)
Retardation Factor
Injection/Extraction Well
Spacing (ft)
Injection Rate (gpm)
Pumping Rate (gpm)
Operating Life of ISL Well
Pattern (yrs)
Time until Peak Arrival (days)
Relative Peak Concentration
CBD-R1
0.01
20
65
6.5
0.65
1
100
100
10
100
1
5-spot - 100
50
51
3
1.19xl04
5.09X10"1
CBD-R2
0.01
20
65
6.5
0.65
1
100
100
10
25
1
5-spot - 100
50
51
3
4.57xl04
4.73X10'1
2.6.2.10 Summary and Conclusions of Excursion Scenarios
Three excursion scenarios were developed to investigate the potential impacts to nearby
receptors. Forty-nine (49) unique simulations were performed to evaluate excursions within the
mined unit beyond the pumping/injection wells. Potential flow up an abandoned borehole was
investigated with three simulations, and two simulations were conducted to evaluate potential
migration through a discontinuous aquitard.
The sensitivity analysis was focused on those parameters that are most uncertain and for which
the results are most sensitive. These parameters include hydraulic gradients and conductivities,
pumping/injection well spacing, aquifer thickness and size of the discontinuity. All of the
concentrations are calculated as relative concentrations and are adjusted to actual concentrations
during the dose assessment (Chapter 4).
All of the relative peak concentrations versus peak arrival times are shown in Figure 2-18. The
highest relative concentrations are associated with the abandoned borehole, the discontinuous
2-58
-------
confining unit, and Series 6 simulations. The mined interval in the Series 6 simulations is
reduced to a thickness of 20 ft, as compared to 70 ft in the other excursion simulations within the
mined interval.
Several graphs have been constructed that show the effect of well spacing at a hydraulic
conductivity of 100 ft/day and a gradient of 0.001 ft/ft (Figures 2-19 and 2-21). As depicted in
the figures, the relative concentrations are always higher at the larger well spacing because of the
injectant escaping between the wells. This relationship is maintained over a range of hydraulic
conductivities, as shown in Figures 2-22 and 2-23.
Relative concentrations as a function of hydraulic conductivities and gradients are shown in
Figures 2-24 through 2-26. As depicted in the figures, the relative concentration is generally
higher for those simulations with lower hydraulic conductivities and gradients. This relationship
occurs because at the higher concentrations and gradients, the injectant reaches the receptor more
quickly, and at lower concentrations, the injectant mass is distributed over a larger area, which
decreases the concentration within the plume.
A statistical analysis was conducted to assess whether there are correlations among any of the
parameters investigated. The results of this analysis are shown in Figures 2-27 through 2-33 and
demonstrate that there is very little correlation among the parameters tested. In each of these
figures, the relative concentrations are plotted both on a linear and a logarithmic scale. The plots
were initially prepared using a linear concentration scale and exhibited no significant
correlations. However, in numerous runs, the relative concentrations were below 0.1 mg/L. To
expand the delineation of data in this region, the scatter plots were redone using a logarithmic
scale for the ordinate.data. Again, very little correlation between concentration and the various
model parameters was noted.
The higher ratios of advection to dispersion resulted in higher peak concentrations. Furthermore,
an important result regarding the interplay of the local and regional gradients is that the steeper
the regional gradient the narrower the capture zone and the greater the contaminant releases.
2-59
-------
9.00E-01
8.00E-01
7.00E-01
G.OOE-01
5.00E-01
4.00E-01
3.00E-01
Max. Cone.
AB-R3
»
6f
AB-R1 CDB-Rl
-AB-R2
6b6e •• ^ CDB-R2
•Max...
l.OOE+01
l.OOE+02
l.OOE+03
l.OOE+04
l.OOE+05
l.OOE+06
Figure 2-18: Maximum Relative Concentrations versus Time for All Excursion Simulations
2-60
-------
01
n n
-------
Effect of Well Spacing on Concentration - K=100, i = 0.1 and
0.001
Run 2b- 50 fl
•Run Ib- 250 Fl
Run 2c, 50 fl
•Run Ic, 250 fl
1000 2000 3000
Distance (ft)
4000
Figure 2-21: Effect of Well Spacing at Constant Hydraulic Conductivity of 100 ft/day and
Hydraulic Gradients of 0.01(Runs Ib, 2b) and 0.001(Runs Ic, 2c).
Effect of Hydraulic Conductivity on Concentration for Well
Spacingsof 50 and 250ft (i = 0.1) Receptor well at 528 ft
0.25
•50ft
•250ft
100
hydraulic conductivity - ft/day
Figure 2-22: Effect of Well Spacing on Concentration as a Function of Hydraulic
Conductivity at a Constant Hydraulic Gradient of 0.1.
2-62
-------
Effect of Hydraulic Conductivity on Concentration for Well Spacings
of 50 and 250ft (i = 0.01) Receptor well at 528ft
o.i
.2 0.08
0.06
c
o
u
O)
>
0.04
0.02
•50ft
•250ft
100
hydraulic conductivity - ft/day
Figure 2-23: Effect of Well Spacing on Concentration as a Function of Hydraulic
Conductivity at a Constant Hydraulic Gradient of 0.01.
0.25
0.2
0.15
0.1
0.05
—»— Run Ig, K=l, i=0.01
-•—Run Ih, K=10, i=0.01
-*—Run li, K=100, i=0.01
-*— Run Id, K=l, i=0.1
-*—Run Ic, K 10, i 0.1
RLIII IF, K 100, i 0.1
1000
2000
3000
4000
Figure 2-24: Relative Concentrations as a Function of Hydraulic Conductivities
and Gradients
2-63
-------
0,25
0.2
0,15
0.1
0,05
•Run Ig, K=l, i=0.01
•Run Id, K=l, i=0.1
1000
2000
3000
4000
Figure 2-25: Relative Concentrations as a Function of Hydraulic Gradient
(Run Id - 0.1, Run Ig - 0.01) and Hydraulic Conductivity of 1 ft/day
0.07
0.06
0.05
0.04
0.03
0.02
0.01
•Run li, K=100, i=0.01
•Run If, K=100, i=0.1
1000
2000
3000
4000
Figure 2-26: Relative Concentrations as a Function of Hydraulic Gradient
(Run If- 0.1, Run li - 0.01) and Hydraulic Conductivity of 100 ft/day
2-64
-------
u. /
n R
u.o
e
CO
s n *=;
kf) U.U
15
E n 4
.2 w.*r
IS
^
= n ^ -
u
C
o n 9 -
X
= 01
n .
t
*
y = 0.0003x + 0.0242
$ R2 = 0.0933
,
* _—"""
*— — • ""
r-f"""! t
0
200 400
Pumping Rate (gpm)
600
Max. Concentration at 528 ft
i t^u •
1 F 1
1 C- 1
•\ p o
1 t-Z
1 p ^
I C-O
1 F 4
•1 p c
1 t-D
i c « _
. : :
•
* *
1^
*
- "^ *
-"^' .
y = 0.0006e°0125x
R2 = 0.0772
' « .
*
:
0
200 400
Pumping Rate (gpm)
600
Figure 2-27: Correlation of Relative Concentration to Pumping Rate
2-65
-------
u. /
n R
•w
CO
^ n ^
u^ U.Q
73
S n 4
° U.H
!
+*
c n ^
o
0
° 0 2 -
X
i
n -\
in
0
*
i
y = -0.0622x +0.0766
R2 = 0.0004 •
» A
|_ *
*^^ ^»
t !
0
0.05 0.1
Hydraulic Gradient
0.15
I C^U
s1D1
CM
un
c
o
(1)
0
£ 1P 4
O I t-H
0
re
S-( p c:
I t-D
»
^-"'^
» i
K
!
_,-J
V = 0.001 2e32782>;
R2 = 0.0587
:
0
0.05 0.1
Hydraulic Gradient
0.15
Figure 2-28: Correlation of Relative Concentration to Hydraulic Gradient
2-66
-------
U. 1
n R
•u
CO
fM n c
m U.O
o 0 A.
.2 U.H
"£ n ^
4) U.O
0
o
° 0 2 -
X
0*1
0
u
c
1 F+f
I CTl.
1 F 1
4= "=-
CO
(N
if)
*• 1 F '
BJ I C-i
0
5 i p <-
t 't'1-
5
X
(0
S-1 F £
1 t-*.
IF.f
»
>
^
|
)
>
>
)
•
:
*
*
:
y = 0.0001x + 0.0689
R2 = 0.0015
. *
9
f $
i
so 100 1;
Hydraulic Conductivity (ft/day)
1
. *
1 1
> t ^
. --" *
^-" *
f-—"'^
y = 0.0007e°037x
« R2 = 0.0862
t
•
50
0
50 100
Hydraulic Conductivity (ft/day)
150
Figure 2-29: Correlation of Relative Concentration to Hydraulic Conductivity (ft/day)
2-67
-------
Max. Concentration at 528 ft
u
0
0
0
0
0
0
/
R
c
Q
o
*
t
y = -0.007x +0.0856
P R2 = 0.0255
1 »_ *
nil •
0 5 10
Hydraulic Gradient times Hydraulic Conductivity
15
Max. Concentration at 528 ft
1 F 1 4
1F 9 -
-i p o <
I C-G <
'I F 4
1 1-4
1 E 5
1
fc * _— = — ~t
_____ — - — -
y = 0.0028e°158x
R2 = 0.0077
0 5 10 15
Hydraulic Gradient times Hydraulic Conductivity
Figure 2-30: Correlation of Relative Concentration to Hydraulic Gradient times
Hydraulic Conductivity
2-68
-------
u. /
OR
CO
01 n ^
ui U.O
£ n A
o U.4
1
"£ n ^ -
4) U.O
u
o n 9
X
^ n 1
n .
*
*
J
0
y = 0.0002x + 0.0455
R2 = 0.0104 *
* .
w *
f T :
0
100 200
Well Spacing
300
1E+0
~ 1E-1
CO
o
' 1E-3
§ 1E-4
u
I 1E-5
1E-6
i
y = 0.0012e°'0075x
R2
= u.ouyb
$
tT
I
^. — — — '
•
* *
:
0
100 200
Well Spacing
300
Figure 2-31: Correlation of Relative Concentration to Well Spacing
2-69
-------
0.7
0.6
OQ
S 0.5 -
ta
o 0.4
y = -0.0049x +0.1145
0
k
R2 = 0.0508
I I I I I 1 I I I I I I I I T
5678 910111213141516171819202122232425
Pumping Array
1E+0
_ 1E-1
U—
CO
IN
10 -IPO
is 1b-2
5
'•S -\ p <5 A
(0 I C-O o
E
-------
OG
CO
u
° n 9 -
w u.z
X
(5
5 n 1 -
U. I
n -
|
*
y = 0.1281x- 0.0902
R2=0.1649
t
t-^'
^.^
LI
^^
*
0
1 2
Layer Model
Max. Concentration at 528 ft
1 t^U
1 p "1
1 t- 1
IPO
1 P-Z
•i p o _
1 P-G
1
1 P 4
I P-H
1 p c _
1 C-O
1P.R -
i
I
•,
^
* -— _
= 0.0131e-°997><
R2 = 0.006 ^
t
*
:
0
1 2
Layer Model
Figure 2-33: Correlation of Relative Concentration to Model Layer
2-71
-------
3.0 PATHWAY DOSE AND RISK CONVERSION FACTORS
To calculate doses and risks to individuals who use water from a well that is contaminated with
radioactivity from the ISL facility, pathway dose and risk conversion factors (PDCFs and
o
PRCFs) were developed and used. PDCFs/PRCFs are defined as the dose/risk received by an
individual (e.g., millirem/year or latent cancer fatality [LCF]/year) divided by the radionuclide
concentration in the well water (e.g., pCi/m3). Because of linearity, the annual dose/risk can be
calculated by multiplying the calculated radionuclide concentration in ground water by the
corresponding PDCF/PRCF. PDCFs/PRCFs include dose and risk coefficients due to the intake
of radionuclides into the human body. The radionuclide-specific dose and risk coefficients used
in calculating the PDCFs/PRCFs were obtained from Federal Guidance Report No. 13 (FOR 13)
(Eckerman et al. 1999 and EPA 2002; see Section 3.2.2 of this report). PDCFs/PRCFs also
include factors that describe the movement and uptake of the radionuclide within the biosphere
(e.g., irrigation rates, plant and animal bioaccumulation factors, weathering and other removal
mechanisms, etc.), as well as human ingestion rates. As described in Section 3.2, for this
analysis, ranges of values were selected for each of these factors and the PDCF/PRCF model was
implemented in spreadsheet form to allow calculation of doses via the ingestion pathway from
the use of well water for drinking and irrigation. The total PDCFs/PRCFs are composed of the
following individual ingestion pathways:
• Ingestion of Drinking Water
• Inadvertent Ingestion of Soil
• Ingestion of Vegetables
• Ingestion of Milk
• Inge sti on of Meat
The general equations for calculating the total PDCFs/PRCFs are shown below:
PDCFTotal = ^PDCFP = ^DClng Ingp f(b,r)P
Ingp f(b,r)P
P P 3-1
where DCing and RDing are the dose and risk coefficients from FOR 13, Ingp is the exposure
pathway human ingestion rate, andf(b,r)p is a pathway-specific function that describes the
buildup (b) and removal (r) (if any) of the radionuclide in water, soil, vegetables, milk, or meat.
For example, the direct consumption of well water, f(b,r)P is set to 1.
Figure 3-1 shows the exposure pathways that were considered in this analysis, while Section 3.1
describes the specific mathematical equations (including the form off(b,r)p) for each exposure
pathway that was analyzed in detail. In addition to the pathways that were analyzed in detail,
there were a number of exposure pathways that were considered, but not analyzed, these are
discussed in Section 3.5 and are shown in grey in Figure 3-1.
' Pathway dose/risk conversion factors are sometimes referred to as biosphere dose/risk conversion factors.
3-1
-------
Figure 3-1: Exposure Pathways Analyzed
Figure 3-1 also shows that there are three sources of radioactivity at an ISL facility: the Plant,
where uranium is removed from the lixiviant; the Lixiviant; and Evaporation (or Holding) Ponds.
This analysis only considers exposures due to the Lixiviant source, because (as indicated in
Figure 3-1) the Plant is regulated under 10 CFR Part 40, by the Nuclear Regulatory Commission
(or an Agreement State), and Evaporation Ponds are regulated by the EPA under 40 CFR Part 61,
Subpart W. Finally, Figure 3-1 shows that lixiviant surface spills and ground water containing
lixiviant from an excursion could flow to a surface water body. While these pathways exist, they
were not analyzed in this report because the addition dilution provided by the surface water body
would result in lower exposures to an individual than exposures from the ground water to well
pathway.
PDCFs and PRCFs were developed for the following longer-lived radionuclides in the U-238
decay series: U-238, U-234, Th-230, Ra-226, and Pb-210 (see Section 3.2.2).
3-2
-------
3.1 Pathway Dose and Risk Models
The basic mathematical models used to calculate the dose and risk from the ingestion pathways
for this analysis were obtained from the NRC's Regulatory Guide 1 . 109 (NRC 1977). Although
the numerical values for many of the parameters given in Regulatory Guide 1.109 have been
updated in the 30-plus years since its publication, the basic mathematical models remain valid
and form the basis for many of today's computer programs used to calculate radiological
impacts, including CAP88 (Trinity 2007), GENII (PNL 1988), and RESRAD (ANL 2001).
While the Regulatory Guide 1.109 models form the basis for many of today's computer
programs, those computer programs often have refined the models to more accurately reflect
reality. When appropriate, those refined models have been used in this analysis, and are pointed
out in the following discussion.
This section presents the mathematical models that were used to calculate the PDCFs/PRCFs,
while Section 3.2 presents and discusses the values that were assumed for each of the parameters
used in the models.
3.1.1 Ingestion of Drinking Water
The annual effective dose (Ewat,A? nire
from the consumption of unfiltered drinking water is given by:
The annual effective dose (Ewat,A? nirem/yr) and risk (Rwat,A> LCF/yr) to a human of age group A9
3-2
where Cw is the radionuclide concentration in the well from which the water is taken (pCi/m3),
IngwatA is the water consumption rate for an individual of age group A (m3/yr), DCing:A is the
ingestion dose coefficient for age group A (mrem/pCi), and RCing,A,DW is the drinking water risk
coefficient for age group A (LCF/pCi).
3.1.2 Inadvertent Ingestion of Soil
Soil can be inadvertently ingested by humans. The annual effective dose (Es0a,A, mrem/yr) and
risk (Rsoti,A, LCF/yr) to a human of age group A from the inadvertent ingestion of soil is given by:
Soil, A = ,,
R-Soil,A = Cs
where Cs0u is the radionuclide concentration in the soil (pCi/kg), Ings0u,A is the inadvertent soil
ingestion rate for an individual of age group A (kg/yr), DCing:A is the ingestion dose coefficient
9 Age Group A is a generic designator for various age groups for which PDCFs and PRCFs are calculated
as described in detail in Section 3.2.1.
3-3
-------
for age group A (mrem/pCi), and RCing,A,Diet is the dietary ingestion risk coefficient for age group
A (LCF/pCi).10
The radionuclide concentration in the surface soil (Csoii), which is available to an individual for
inadvertent ingestion, is given by:
, = Irr*- - CV 3-4
where Irr is the irrigation rate (m/yr), X is the radionuclide decay constant (yr"1), XR is the
removal rate from soil (yr"1), tb is the period of time soil is irrigated with contaminated water
(yr), and P is the effective "surface density" for soil, (kg/m ).
Although Regulatory Guide 1.109 does not calculate exposures from soil ingestion, equation 3-4
is based on the soil buildup portion of Regulatory Guide 1.109 equation A-8 (NRC 1977). In
addition to radiological decay, equation 3-4 includes removal from the soil by physical processes
(Aft), which is not included in Regulatory Guide 1.109 equation A-8. This term is intended to
model the leaching of radionuclide from the surface layer to soil depths were the radionuclides
are no longer available for uptake by humans, plants, and/or animals. This approach is used in
models such as Peterson (1983), PATHWAY (Whicker and Kirchner 1987), GENII (Napier et al.
1988), Abbott and Rood (1993), RESRAD (Yu et al. 2002), and CAP88 (Trinity 2007).
3.1.3 Ingestion of Vegetables
The annual effective dose (Eveg,A, mrem/yr) and risk (Rveg,A, LCF/yr) to a human of age group A
from the consumption of vegetables grown with unfiltered irrigation water is given by:
EVeg,A = CVeg DDVeg InSveg,A DCIng,A
where Cveg is the radionuclide concentration in vegetables (pCi/kg), Ingcmp,A is the vegetable
ingestion rate for an individual of age group A (kg/yr), DDyeg is the fraction of radioactivity
retained on leafy vegetables and produce after washing, DCing:A is the ingestion dose coefficient
for age group A (mrem/pCi), and RCing,A,Diet is the dietary ingestion risk coefficient for age group
A (LCF/pCi).
According to Regulatory Guide 1.109, the radionuclide concentration in vegetation results from
deposition onto the plant foliage and from uptake from the soil of activity deposited on the
ground. The equation that models the radionuclide concentration in vegetation (Cyeg) is:
10 FOR 13 lists a single dose conversion factor for ingestion of each radionuclide, while separate risk
conversion factors are cited for drinking water (DW) and dietary (Diet) ingestion.
3-4
-------
3-6
where rv is the fraction of deposited activity retained on vegetables (dimensionless), kw is the
removal rate constant for physical removal by weathering (yr"1), tv is the length of the vegetable
growing season (yrs), Yv is the vegetable crop productivity or yield (kg/m2), Bv is the element-
specific vegetable uptake factor from soil (pCi/kg vegetables per pCi/kg soil), and B'v represents
the net effect of all resuspension processes (NCRP 1999).
In equation 3-6, the first term in brackets relates to the concentration derived from direct foliar
deposition during the growing season, while the second term relates to root uptake of
radionuclide contamination from soil and reflects the long-term deposition during operation of
the uranium recovery facility.
Equation 3-6 is based on Regulatory Guide 1.109, equation A-8 (NRC 1977), except that two
additional features have been included in the root uptake term. First, in addition to radiological
decay, equation 3-6 includes removal due to removal from the soil (A#), which was described in
Section 3.1.2. Second, as stated by the National Council on Radiation Protection and
Measurements (NCRP) in Report No. 129 (NCRP 1999, page 92):
In the case of contamination by resuspended soil, the mechanisms can be quite
complex. Not only can airborne resuspension and subsequent redeposit
contaminate the vegetation, but also phenomena such as rain splash, saltation
and mechanical disturbances during harvest. Thus, B'v must also be determined
empirically for the particular site or type of site and the type of vegetation.
To include this effect, Regulatory Guide 1.109, equation A-8 has been modified to include a
transfer factor representing the net effect of all resuspension processes (B'v\ as recommended by
the NCRP Report No. 129, equation 5.2 (NCRP 1999).
3.1.4 Ingestion of Milk
The annual effective dose (Euk,A-, mrem/yr) and risk (Ruk,A, LCF/yr) to a human of age group A
from the consumption of milk from a cow that ingests contaminated well water, fodder grown
with unfiltered irrigation water, and contaminated soil is given by:
where Cm is the radionuclide concentration in milk (pCi/L) and IngMk,A is the individual milk
consumption rate (L/yr). The CM/C term is calculated using the following equation:
(lngMkw Cw + IngMk^F CFod + Inguk^oll CSoil ) 3-8
3-5
-------
where CFuk is the element-specific cow's intake to milk transfer coefficient (pCi/L milk per
pCi/day intake), Cp0d is the radionuclide concentration in the animal's fodder (pCi/kg) (see
equation 3-9), IngMkfod is the consumption rate of fodder by the cow (kg fresh weight of
fodder/day), IngMk,w is the consumption rate of water by the cow (m3/day), CSM is the
radionuclide concentration in the soil (Bq/kg) (see equation 3-4), and IngMk,Soii is the
consumption rate of soil by the cow (kg/day). The radionuclide concentration in animal fodder is
given by:
CFod=Irr
rf
Cw 3-9
where r/is the fraction of deposited activity retained on fodder (dimensionless), Xw is the removal
rate constant for physical removal by weathering (yr"1), t/is the length of the fodder growing
season (yrs), 1/is the fodder productivity or yield (kg/m ), 5/is the element-specific
concentration fodder uptake factor from soil (pCi/kg fodder / pCi/kg soil), and B'f represents the
net effect of all resuspension processes (NCRP 1999).
3.1.5 Ingestion of Meat
The annual effective dose (EMI.A-, mrem/yr) and risk (RutA, LCF/yr) to a human of age group A
from the consumption of meat from cattle that ingests contaminated well water, fodder grown
with unfiltered irrigation water, and contaminated soil is given by:
EMt,A = CMt In§Mt,A DCIng,A
R-Mt,A = ^Mt *nSMt,A R^- Ing,A,Diet
where CM is the radionuclide concentration in meat (pCi/kg) and Ingut,A is the individual
consumption rate of the animal product (kg/yr). The Cut term is calculated using the following
equation:
CMt = CFMt lngmw Cw + IngMtf CFod + IngMtiSoil CSoil 3-11
where CFut is the element-specific cattle intake to meat transfer coefficient (pCi/kg meat per
pCi/day intake), Cp0d is the radionuclide concentration in the animal's fodder (pCi/kg fresh
weight of fodder) (see equation 3-9), Ingut,Fod is the consumption rate of fodder by cattle (kg
fresh weight of fodder/day), IngMt,w'^ the consumption rate of water by cattle (m /day), Cs0n is
the radionuclide concentration in the soil (pCi/kg) (see equation 3-4), and Ingut.Soii is the
consumption rate of soil by cattle (kg/day).
3.1.6 Pathway Dose and Risk Factors
The pathway dose and risk factors are calculated by dividing the above derived exposure
equations by the radionuclide concentration in the well from which the water is taken and then
summing over all of the exposure pathways, as shown below:
3-6
-------
where P identifies the exposure pathway (i.e., drinking water, soil ingestion, vegetable ingestion,
ingestion of milk, and ingestion of meat), and^4 identifies the age group (i.e., infant, child, teen,
and adult).
3.1.7 Implementation
In order to calculate the PDCFs and PRCFs, the above equations were programmed into an
Excel® spreadsheet. To propagate the uncertainty in the value for many of the parameters used
in the above equations, the Excel add-in Crystal Ball was used to solve for the PDCFs and
PRCFs using a Monte Carlo simulation. Parameter uncertainty is due to both natural variation
between individuals (e.g., drinking water, vegetable, milk, and meat consumption rates \IngDW,A,
Ingveg,A, IngMk,A, and IngMt^\, etc.) and uncertainty as to the "true" value for the parameter (e.g.,
vegetable, fodder, milk, and meat transfer coefficients \BV, Bf, CFmiik, and CFmeat], etc.). Instead
of selecting a single value to represent each parameter, the Crystal Ball Monte Carlo simulation
will randomly sample from a range of values for each parameter and solve for the PDCFs and
PRCFs, and then it will repeat the entire process until it has calculated a range of PDCFs and
PRCFs. From this range of values, mean, median, 90* percentile, and other measures can be
selected. For this analysis, the Monte Carlo simulation repeated the calculations 1,000 times.
3.2 Input Parameters
This section describes how the values for the various input parameters were selected and
identifies those parameters for which parameter distributions were specified.
3.2.1 Age Groups
As shown above in Section 3.1 and in keeping with Regulatory Guide 1.109, the PDCFs and
PRCFs have been calculated for four age groups—infant, child, teen, and adult. In order to
calculate the age-specific PDCFs and PRCFs, it was necessary to use age-specific dose and risk
conversion factors and exposure factors (e.g., water and food consumption rates). Fortunately,
the source documents for these parameters present age-dependent data. Unfortunately, the
manner in which the data are grouped by age is not uniform across each source document, or
even for different parameters within a single source document.
The age-dependent parameters used in this analysis were obtained from two primary sources:
Federal Guidance Report No. 13 CD Supplement (EPA 2002) for dose and risk coefficients, and
the Exposure Factors Handbook (EFH) (EPA 201 Ib) for usage factors.
3-7
-------
The dose and risk coefficients used in this report were developed based on the recommendations
of the International Commission on Radiological Protection (ICRP). Starting in the mid 1980s,
the ICRP began developing age-dependent dose coefficients for members of the public (ICRP
2006). For example, in Publication 72 (ICRP 1995), the ICRP has indicated that the following
six age-specific dose coefficients are appropriate for calculating doses for the indicated age
ranges:
ICRP Age Group Applicable Age Range
3 months from 0 to 1 year
1 year from 1 year to 2 years
5 year more than 2 years to 7 years
10 year more than 7 years to 12 years
15 year more than 12 years to 17 years
Adult more than 17 years
However, in Publication 101 (ICRP 2006), the ICRP recommended:
... that the annual dose for the representative person should be defined by three age
categories. These categories are 0-5years (infant), 6-15 years (child), and 16-70years
(adult). The shorter time period is selected for the infant age category, when dosimetric
characteristics are changing most rapidly, to avoid any unwarranted reduction in the
importance attached to doses to younger age groups. Use of these three age categories is
sufficient to characterize the radiological impact of a source and to ensure consideration
of younger, more sensitive populations. For practical implementation of this
recommendation, dose coefficients and habit data for a 1-year-old (infant), a 10-year-old
(child), and an adult should be used to represent the three age categories. (ICRP 2006)
The age grouping recommendations from ICRP Publication 101 are summarized in the following
table.
Age category (years) Name of age category Dose coefficient and habit data to be used
0-5 Infant 1 year old
6-15 Child 10 year old
16-70 Adult Adult
Unlike the dose and risk coefficients, the exposure factors used in this study were taken from the
EPA's EFH (EPA 201 Ib). Many of the exposure factors presented in the Handbook are broken
down by age, and the age grouping (at least for children) is that recommended by the EPA (EPA
2005). EPA 2005, Table 4, recommends the following 10 age groups for children under the age
of 21 years:
Children <1 Year Children >1 Year
Birth to <1 month 1 to <2 years
1 to <3 months 2 to <3 years
3 to <6 months 3 to <6 years
6 to < 12 months 6to
-------
Table 3-1 presents the age groups for the various parameters from each source document mapped
into the four age groups used in this analysis. The four age groups used in this analysis assume
that the population is made up of infants, children, teenagers, and adults (NRC 1977). In Table
3-1, the parameter (or distribution) values used in this analysis are shown in bold.
Table 3-1: Age Groups Used in the Analysis
Age
Group
Infant
Child
Teen
Adult
FGR 13*
Risk
Coefficients
0 to 5 yr**
5 to 15
15 to 25
25 to 70
Dose
Coefficients
100 days
lyr
5
10
15
20
Exposure Factors Handbook*
Water
Consumption
Birth to <1 mo
1 to <3 mos
3 to <6 mos
6 to <12 mos
lto<2
2to<3
3to<6
6to21
>65
Soil and Dust
6 wks to <1 yr
lto<6
3to<6
6 to 21
6 to 21
Vegetables,
Milk and
Meat
Birth to 1 yr
Ito2
3 to 5
6 to 12
13 to 19
20 to 49
>50
Body Weight
Birth to <1 mo
1 to <3 mos
3 to <6 mos
6 to <12 mos
lto<2
2to<3
3to<6
6to80
* Items in bold were selected for use in this analysis.
** For Child, Teen, and Adult, all ages are given in years.
Although different groupings could be made, based on the analysis performed by the ICRP
(ICRP 2006), it is not anticipated that any alternative age grouping would have a significant
effect on the calculated doses and risks.
3.2.2 Dose and Risk Conversion Coefficients
Naturally occurring uranium found in the ground contains (by weight) 99.3% U-238, 0.7%
U-235, and a trace amount of U-234. In terms of the amount of radioactivity, natural uranium
contains approximately 48.6 % U-238, and 49.2 % U-234, and 2.2 % U-235. Because U-235
composes such a small portion of the radioactivity in natural uranium and because the U-235
ingestion dose and risk conversion coefficients are very similar to the U-234 and U-238
coefficients, U-235 has not been included in this analysis. In addition to being radioactive,
uranium is hazardous from the standpoint of chemical toxicity. The main chemical effect
associated with exposure to uranium is irreversible kidney damage. Although this section focuses
on uranium's radioactive toxicity, its chemical toxicity risk is discussed in Section 4.5, along
with the chemical toxicity risk of various other metals found in the ore.
As shown in Figure 3-2, the uranium radioactive decay series contains five radionuclides with
half-lives of over a year. They are U-238 (4.5xl09 years), U-234 (2.4xl05 years), Th-230
(7.7xl04 years), Ra-226 (1,600 years), and Pb-210 (22.3 years). This analysis focuses on these
3-9
-------
five radionuclides. If other shorter-lived radionuclides are initially present in the lixiviant, they
will decay before they reach the receptor well. However, the longer-lived radionuclides will
decay as they travel to the receptor well, so the shorter-lived radionuclides will be present at the
well. This analysis addresses the in-growth of the shorter-lived radionuclides by including the
dose and risk coefficients of the short-lived progeny with the parent's coefficients. For example,
the Ra-226 ingestion dose and risk coefficients used in this analysis include the contributions
from Po-210, Bi-210, Pb-210, Bi-214, and Pb-214 [the FOR 13 CD Supplement (EPA 2002)
does not provide ingestion dose or risk factors for Pa-234m, Rn-222, Po-218, or Po-214, likely
because Rn-222 is a noble gas, and the two polonium isotopes and protactinium half-lives are too
short for them to enter the food chain, as shown in Figure 3-2].
U-238 .
4.5x1093
U-234*
2.4x10*3
Pa-234ffi
1.17m
S
• Th-234
24.1 d
*
/
Th 230
7.7x10* a
Elements Names
U - uranium
Th thorium
Ra radium
Pa protactinium
Rn radon
Po polonium
Bi = bismuth
Pb = lead
- 3lphs decay
/
- beta decay
Only m3in decsys 3re shown
Gamma emitters are not indicated
Ra 226
1600 a
3.82 d
Half-life units
a - )
d -
S -
/ears
days
lours
ninutes
seconds
. Po-218 .
3.05m
Po-214^
1.6x10^5
If
138.4d
Bi-214^
19.9m
/
If
Bi-210
5.0 d
/
. Pb-214
26.8m
. Pb-210
22.3 a
Stable
Figure 3-2: Uranium Decay Series
The dose and risk coefficients used in this analysis were obtained from FGR 13 (Eckerman et al.
1999). FGR 13 does not actually present age-specific risk coefficients or any dose coefficients.
Instead, it presents risk coefficients that are representative of the U.S. population. However, in
order to calculate the FGR 13 risk coefficients, it was necessary to calculate age-specific dose
and risk coefficients. Those age-specific dose and risk coefficients are included in data files
supplied with the FGR 13 CD Supplement (EPA 2002):11 FGR13ING.GBD contains the dose
coefficients due to ingestion of radioactive material and FRG13ING.RBS contains both the
drinking water and dietary ingestion risk coefficients. The dose conversion factors use the latest
methods and models from the ICRP and are analogous to the dose factors contained in ICRP
Publication 72 (ICRP 1995). The dose and risk coefficients from FGR 13 that were used in this
FGR13ING.GBD and FGR13ING.RBS, containing age-dependent ingestion dose and risk factors, are also
supplied with the CAP88 and GENII computer programs.
3-10
-------
analysis are shown in Table 3-2. In Table 3-2, the Ra-226 ingestion dose and risk factors that
include the contribution from the short-lived progeny are shown as Ra-226+P.
Table 3-2: Radionuclide-Specific Ingestion Dose
and Risk Coefficients
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Mortality Coefficient - Drinking (LC/Sv)
Infant
0-5
2.56E-07
1.71E-09
1.25E-07
3.83E-07
3.55E-11
6.51E-11
3.92E-08
4.22E-07
1.27E-08
9.99E-09
4.48E-09
9.06E-09
1.35E-08
Child
5-15
1.02E-07
6.93E-10
6.92E-08
1.72E-07
1.62E-11
2.80E-11
2.59E-08
1.98E-07
5.02E-09
4.59E-09
1.81E-09
4.16E-09
5.97E-09
Teen
15-25
4.58E-08
1.92E-10
3.08E-08
7.68E-08
9.51E-12
1.34E-11
1.97E-08
9.65E-08
2.73E-09
2.05E-09
5.19E-10
1.86E-09
2.38E-09
Adult
25-70
2.47E-08
3.30E-11
8.78E-09
3.35E-08
1.86E-12
2.61E-12
3.07E-09
3.66E-08
1.07E-09
6.60E-10
7.78E-11
6.05E-10
6.83E-10
Mortality Coefficient - Dietary (LCF/Sv)
Infant
0-5
2.52E-07
1.71E-09
1.24E-07
3.78E-07
3.54E-11
6.48E-11
3.86E-08
4.16E-07
1.20E-08
9.94E-09
4.49E-09
9.01E-09
1.35E-08
Child
5-15
1.03E-07
6.95E-10
6.93E-08
1.73E-07
1.62E-11
2.80E-11
2.59E-08
1.99E-07
5.03E-09
4.60E-09
1.81E-09
4.17E-09
5.98E-09
Teen
15-25
4.6E-08
1.95E-10
3.10E-08
7.72E-08
9.5E-12
1.34E-11
1.99E-08
9.71E-08
2.73E-09
2.06E-09
5.25E-10
1.87E-09
2.40E-09
Adult
25-70
2.61E-08
3.74E-11
9.38E-09
3.55E-08
2.18E-12
3.00E-12
3.31E-09
3.88E-08
1.18E-09
7.30E-10
8.90E-11
6.68E-10
7.57E-10
3-11
-------
Table 3-2: Radionuclide-Specific Ingestion Dose
and Risk Coefficients
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Effective Dose Coefficient (Sv/Bq)
Infant
100 days
2.60E-05
1.50E-08
8.31E-06
3.43E-05
1.37E-09
2.17E-09
4.65E-06
3.90E-05
4.13E-06
3.69E-07
3.99E-08
3.34E-07
3.74E-07
Child
5 yrs
4.37E-06
4.84E-09
2.18E-06
6.55E-06
3.66E-10
5.21E-10
6.16E-07
7.18E-06
3.09E-07
8.84E-08
1.26E-08
8.01E-08
9.27E-08
Teen
15 yrs
1.57E-06
1.63E-09
1.92E-06
3.49E-06
1.42E-10
2.03E-10
1.52E-06
5.01E-06
2.19E-07
7.45E-08
4.23E-09
6.71E-08
7.13E-08
Adult
20 yrs
.21E-06
.31E-09
6.96E-07
.91E-06
.12E-10
.39E-10
2.80E-07
2.19E-06
2.14E-07
4.95E-08
3.40E-09
4.45E-08
4.79E-08
Source: FOR 13 CD Supplement (EPA 2002)
The dose factors from Table 3-2 were converted to the dose coefficients for ingestion
(DCingtA, mrem/pCi) by multiplying by 0.037xl05 (Bq/pCi)(mrem/Sv), while the risk factors
from Table 3-2 were converted to the risk coefficients for ingestion (RCing,A,Dw and RCing,A,Diet,
LCF/pCi) by multiplying by 0.037 (Bq/pCi).
Mortality Versus Morbidity Risk Coefficients
For risks due to non-radiological sources, the EPA typically bases its rulemaking on cancer
morbidity risk, e.g., the National Contingency Plan (55 Federal Register 8665-8865, March 8,
1990). However, for risks due to radiological sources, the EPA has traditionally used cancer
mortality as the basis for their rulemaking, e.g., the Radiation Protection Standards for Yucca
Mountain (73 Federal Register 61256- 61289, October 15, 2008). A brief comparison of cancer
morbidity to mortality risks is presented here.
Based on analysis by the EPA (EPA 1999a), the Interagency Steering Committee on Radiation
Standards (ISCORS 2002) recommended a cancer morbidity risk coefficient of 8x 10"4 per rem
(8x 10"6 per sievert) and a cancer mortality risk coefficient of 6x 10"4 per rem (6x 10"6 per sievert).
Thus, the morbidity to mortality risk ratio is 1.33. Alternatively, the National Academy of
Sciences uses a ratio of 1.5 for total cancer incidence to fatal cancer incidence. Depending upon
exposure pathways, radionuclide, total inventory and site characteristics, the ratio of 1.5 could
be off by a factor of 2. (NAS 1995, page 51)
Finally, Table 3-3 shows the morbidity to mortality risk ratio based on the FOR 13 drinking
water risk coefficients. As Table 3-3 shows, the ratio ranges from 1.30 for the adult Ra-226+P
risk coefficients to 1.70 for the infant U-238+P risk coefficients.
3-12
-------
Table 3-3: FGR 13 Drinking Water Cancer
Morbidity / Mortality Risk Ratio
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Morbidity / Mortality Risk Ratio
Infant
0-5
1.43
1.81
1.42
1.43
1.23
1.44
1.55
1.44
1.56
1.64
1.81
1.64
1.70
Child
5-15
1.42
1.80
1.41
1.42
1.23
1.42
1.49
1.43
1.56
1.61
1.81
1.61
1.67
Teen
15-25
1.38
1.80
1.36
1.37
1.17
1.30
1.42
1.38
1.46
1.53
1.80
1.52
1.58
Adult
25-70
1.30
1.77
1.30
1.30
1.17
1.29
1.38
1.31
1.39
1.44
1.80
1.43
1.47
Dose/Risk Coefficient Uncertainty
In FGR 13, the dose and risk coefficients are presented as single values, rather than as a
distribution or range of values. However, in order to calculate the FGR 13 dose and risk
coefficients, numerous assumptions had to be made regarding biokinetic, dosimetric, and
radiogenic cancer risk models. Different values could have been assumed regarding these models
and/or parameter values, which would have resulted in different values for the dose and risk
coefficients. FGR 13, Table 2.4 and Appendix D, briefly describe the uncertainty in the dose and
risk coefficients that result from these assumptions. Pawel et al. (2007) provide additional insight
into the uncertainty associated with the FGR 13 risk coefficients. As Pawel et al. (2007)
explains:
Assigned levels of uncertainty were based on sensitivity analyses in which various
combinations of plausible biokinetic and dosimetric models and radiogenic
cancer risk models were used to generate alternative risk coefficients.
Uncertainties relating to the validity of the linear-no-threshold hypothesis were
not addressed in the analysis because this is not feasible.
The uncertainty in a risk coefficient was viewed as the net result of uncertainties
in the following main components of the derivation: biokinetic models describing
the biological behavior of ingested or inhaled radionuclides; specific energies
that relate emissions from source organs to energy deposition in target organs;
risk model coefficients representing the risk of cancer per unit absorbed dose to
sensitive tissues from low-LET radiation at high dose and high dose rate; tissue-
3-13
-------
specific dose and dose rate effectiveness factor (DDREF); and tissue-specific
high-dose relative biological effectiveness (RBE).
Although this analysis will use the FGR 13 dose and risk coefficients as single values, the
following discussion is given to provide some understanding of the implication of this
assumption.
The results of the Pawel et al. (2007) uncertainty analysis were assigned to one of five
"uncertainty categories" (A through E), depending on the ratio of the 5* to the 95* percentiles of
the calculated risk coefficients. For the five radionuclides of interest in this analysis, Table 3-4
presents the results of the Pawel et al. (2007) uncertainty analysis.
Table 3-4: Risk Coefficient Uncertainty
Nuclide
Pb-210
Ra-226
Th-230
U-234
U-238
Uncertainty Category
C
C
D
C
B
Category Range*
35 < Q95/Q5 < 50
35 < Q95/Q5 < 50
50 < Q95/Q5 < 150
35 < Q95/Q5 < 50
15 < Q95/Q5 < 35
* Q5 and Q95 are the 5% and 95% sample quantiles of the risk
coefficients generated by combining plausible variations of the
biokinetic, dosimetric, and risk models.
Table 3-4 shows that for the five radionuclides of interest, U-238 has the least uncertainty (at
between a factor of 15 to 35), while Th-230 has the most (at between a factor of 50 to 150). For
several radionuclides, Pawel et al. (2007) provides a brief discussion of the basis of the
uncertainty. As it happens, these brief discussions were provided for two of the five
radionuclides of interest for this analysis, and they have been reproduced below.
Regarding the uncertainty of the Ra-226 ingestion risk factors, Pawel et al. (2007) states:
Risk and dose models contribute -90% and 10% of uncertainty, respectively. GI
uptake and systemic biokinetics reasonably well established. Important cancer
sites include bone, for which the risk model is highly uncertain, and colon, for
which risk model is moderately uncertain.
Regarding the uncertainty of the U-234 ingestion risk factors, Pawel et al. (2007) states:
Risk and dose models contribute -90% and 10% of uncertainty, respectively. GI
uptake and most of the important aspects of systemic biokinetics reasonably well
known. Important cancer sites include colon, for which risk model is moderately
well established, and bone, for which risk model is poorly established.
Although it is not possible to incorporate this risk coefficient uncertainty into the present
analysis, the following two examples demonstrate the potential impact of this uncertainty.
3-14
-------
If it is assumed that the FOR 13 U-23 8 dietary risk coefficient of 1.51 x 10"9 Sv"1 is halfway
between the 5th to the 95th percentiles, and that Qgs/Qs is equal to 15, then the average U-238 risk
-9
-1
coefficient has a range from 3.90x 10 to 5.85x 10 Sv" , rather than the FOR 13 value.
Likewise, if it is assumed that the FGR 13 Th-230 dietary risk coefficient of 2.16><10~9 Sv"1 is
halfway between the 5th to the 95th percentiles, and that Qgs/Qs is equal to 150, then the average
Th-230 risk coefficient has a range from 1.76x 10"10 to 2.65x 10"8 Sv"1, rather than the FGR 13
value.
3.2.3 Ingestion of Drinking Water
As shown by equation 3-2, the only parameters necessary to calculate the PDCFwatA and
PRCFwat,A are the age-dependent drinking water consumption rates. The age-dependent drinking
water consumption rates were obtained by multiplying the weight normalized water consumption
rate provided in EFH, Table 3-41 (EPA 201 Ib), and shown in Table 3-5, by the age-specific
body weight, as shown in Table 3-6.
Table 3-5: Age-Dependent Weight Normalized Water
Consumption Rates
Percentile
10%
25%
50%
75%
90%
95%
99%
Water Intake - mL/kg-day
Infant
3 to <6 mos
8
50
95
132
163
186
238
Child
3 to 6 yrs
7
14
24
37
52
63
96
Teen
11 to 16 yrs
4
7
13
20
33
44
66
Adult
>21 yrs
6
12
19
29
41
50
70
Source: EPA 201 Ib, Table 3-41
In order to demonstrate that the water consumption rate data from the EFH was entered into and
used correctly by Crystal Ball, Figure 3-3 shows the distributions of the four age-dependent
water consumption rates calculated by the Crystal Ball Monte Carlo simulation for this analysis
and compares them to the Table 3-5 EFH distributions. As shown, there is very good agreement
between the EFH distributions and the distributions used in this analysis.
3-15
-------
nfant, EFH -—Infant. CB
Child. EFH —Q—Child, CB
Teen, EFH — *~ Teen, CB
Adult, EFH Adult, CB
100 150
WaterConsumption (mL/ kg-day)
Figure 3-3: Distribution of Age-Dependent Weight Normalized
Water Consumption Rates
As indicated above, the daily weight normalized water consumption rates from Table 3-5 were
converted to the individual water ingestion rate (IngWah m3/yr) by multiplying by 365.25* 10~6
(day/yr)(m3/mL), and then multiplying by the individual's body weight. Table 3-6 presents the
age-dependent body weight distributions that were obtained from the EFH, Table 8-3 (EPA
201 Ib).
Table 3-6: Body Weight Distributions
Percentile
5%
10%
15%
25%
50%
75%
85%
90%
95%
Body Weight - kg
Infant
3 to <6 mos
5.7
6.1
6.3
6.7
7.3
8.0
8.4
8.7
9.1
Child
3 to <6 yrs
13.5
14.4
14.9
15.8
17.8
20.3
22.0
23.6
26.2
Teen
11 to <16 yrs
34.0
37.2
40.6
45.0
54.2
65.0
73.0
79.3
88.8
Adult
30 to <40 yrs
53.5
57.4
60.1
66.1
77.9
92.4
101.0
107.0
118.0
Source: EPA 201 Ib, Table 8-3
Figure 3-4 shows the Crystal Ball (CB) calculated combined body weight and drinking water
consumption distributions for all four age groups, and compares them to the Regulatory Guide
1.109, Table E-5 maximum exposed recommended age dependent annual drinking water rates.
3-16
-------
Notice that the R.G. 1.109 recommended maximum rate for adults corresponds to 2 liters per
day, which is commonly used in risk assessments. Also, notice that all of the R.G. 1.109
recommended maximum rates falls within the CB calculated distributions.
Infant, CB
Child, CB
Teen.CB
Adult. CB
--- Infant, R.G. 1.109, Max: 330 (L/yr), 72.3°,(
— Child, R.G. 1.109, Max: 510 (L/yr), 96.3%
--Teen, R.G 1.109, Max: 510 (L/yr), 854%
Adult, R.G. 1.109, Max: 730 (L/yr), 66.3%
0% *
WaterConsumption(L/yr)
Figure 3-4: Distribution of Age-Dependent Water Consumption Rates
Compared to Regulatory Guide 1.109 Recommended Maximum Rates
Figure 3-5 compares the CB calculated combined adult body weight and drinking water
consumption distribution to the drinking water distribution recommended in the RESRAD
documentation (NUREG/CR-6697, Attachment C, Section 5.2). For a cumulative probability of
less than 20%, the RESRAD distribution is larger than the CB distribution, at 50% cumulative
probability the CB value is 33% larger than the RESRAD value, while at 93.3% cumulative
probability the maximum difference of 66% between the CB and RESRAD values was
calculated.
3-17
-------
500 1000 1500
Water-Consumption (L/yr)
Figure 3-5: Distribution of Age-Dependent Water Consumption Rates
Compared to the RESRAD Distribution
Based on Figure 3-4 and Figure 3-5 it is concluded that the Crystal Ball calculated combined
EFH body weight distributions and weight-normalized water consumption rate distributions are
consistent with both Regulatory Guide 1.109 and RESRAD water consumption rates. It is
reasonable to assume that similar results would be obtained for other distributions that result
from the coupling of variables.
3.2.4 Inadvertent Ingestion of Soil
As shown in Section 3.1.2, in order to calculate the PDCFs0n,A and PRCFs0n,A, it is necessary to
know:
• The irrigation rate (Irr)
• The removal rate from soil (Aw)
• The duration period for irrigation (tb)
• The effective "surface density" for soil (P)
• The removal rate constant for physical removal by weathering (Aw)
• Age-dependent inadvertent soil ingestion rates (Ings0u,A)
The irrigation rate (Irr) was obtained from the 2002 Census of Agriculture Farm and Ranch
Irrigation Survey (USDA 2004), which presented the area of farmland irrigated and the volume
of irrigation water used for each state. Irrigation rates (m/yr) were calculated for each state, and
the cumulative distribution shown in Figure 3-6 was developed and used in this analysis. As a
check of the 2002 census data, NUREG/CR-5512 (SNL 1999) was consulted. Table 6.18 of
NUREG/CR-5512, Volume 3, contains data from the 1992 Census of Agriculture for 27 states.
3-18
-------
Figure 3-6 contains the cumulative distribution of the NUREG/CR-5512 data, which shows very
good agreement with the 2002 Census of Agriculture cumulative distribution.
2002 Census of Agriculture Farm and
Ranch Irrigation Survey
NUREG/CR-5512, Volume 3, Table 6.18
0.75 1 1.25
Irrigation Rate (m/yr)
Figure 3-6: Distribution of Irrigation Water Application Rate
The removal rate from soil (}IR) and the duration period for irrigation (tb) were taken from the
CAP88 Users Manual (Trinity 2007) as a constant 0.02 (yr"1) and 100 yrs, respectively. Based on
NUREG/CR-6697, Appendix C, Section 3.1 (ANL 2000), the effective "surface density" for soil
(P) was assumed to have a normal distribution, with a mean and standard deviation of 1.52 and
0.23 g/cm3, and minimum and maximum values of 0.83 and 2.21 g/cm3.
The inadvertent soil ingestion rates were modeled as a triangular distribution, with minimum,
most likely, and maximum values for each age group, as shown in Table 3-7. The most likely
soil ingestion rates in Table 3-7 were obtained from the general population central tendency
values in EFH (EPA 201 Ib), Table 5-1 for soil plus dust ingestion. Likewise, the maximum
Child soil ingestion rate was obtained from the child general population upper percentile value in
EFH (EPA 201 Ib), Table 5-1 for soil plus dust ingestion. The minimum and the other age group
maximum values were assumed.
Table 3-7: Soil + Dust Ingestion
Age Group
Infant
Child
Teen
Adult
Soil Ingestion Rate (nig/day)
Minimum
0
0
0
0
Most Likely*
60
100
50
50
Maximum
200
200*
200
200
EPA 20lib, Table 5-1
3-19
-------
3.2.5 Ingestion of Vegetables
As shown in Section 3.1.3, in order to calculate the PDCFveg,A and PRCFveg,A, it is necessary to
know:
• The fraction of radioactivity retained on leafy vegetables and produce after washing
(DDVeg)
• The fraction of deposited activity retained on vegetables (rv)
• The removal rate constant for physical removal by weathering (Aw)
• The length of the vegetable growing season (/„)
• The vegetable productivity or yield (Yv)
• The element-specific vegetable uptake factor from soil (6V)
• The net effect of all resuspension processes (B'v)
• Age-dependent vegetable ingestion rates (Ingveg,A)
The following terms have already been defined in Section 3.2.4:
• The irrigation rate (Irr)
• The removal rate from soil (Aw)
• The duration period for irrigation (Y/,)
• The effective "surface density" for soil (P)
• The removal rate constant for physical removal by weathering (Aw)
The geometric means and standard deviations of the lognormal distributions for the element-
specific vegetable uptake factor from soil (6V) used in this analysis were taken from NCRP 1999,
and are shown in Table 3-8.
Table 3-8: Element-Specific Input Parameters for Lognormal Distribution
Element
Pb
Ra
Th
U
Bv
Vegetable Uptake
(pCi/g vegetables
per pCi/g soil)
GM
4E-03
0.04
1E-03
2E-03
GSD
2.5
2.5
2.5
2.5
Bf
Fodder Uptake
(pCi/g fodder per
pCi/g soil)
GM
0.09
0.2
1E-03
0.01
GSD
2.5
2.5
2.5
2.5
CFM,
Meat Transfer
(pCi/kg meat per
pCi/day intake)
GM
8E-04
1E-03
1E-04
8E-04
GSD
2.0
2.0
2.8
2.0
CF]y[k
Milk Transfer
(pCi/L milk per
pCi/day intake)
GM
3E-04
1E-03
5E-06
4E-04
GSD
2.5
1.6
2.5
1.8
Note: GM is the geometric mean, and GSD is the geometric standard deviation
Source: NCRP (1999)
The fraction of deposited activity retained on vegetables (rv) was assumed to be a uniform
distribution with a minimum of 0.1 and a maximum of 0.6, as recommended for leafy and other
vegetables by NUREG/CR-5512, Volume 3, Table 6.76 (SNL 1999). The fraction of
radioactivity retained on leafy vegetables and produce after washing (DDVeg) was taken as 0.5, as
specified in Section 10.2 of the CAP88 Users Manual (Trinity 2007).
3-20
-------
The removal rate constant for physical removal by weathering (Aw) was assumed to be a
triangular distribution with minimum, maximum, and most likely values of 5.1, 84, and 18 yr"1,
respectively, as recommended by NUREG/CR-6697, Appendix C, Section 6.6 (ANL 2000). It is
noted that the most likely value has a weathering half-life of 14 days, which is identical to the
value recommended in CAP88, Section 10 (Trinity 2007); Regulatory Guide 1.109, Table E-15
(NRC 1977); andNUREG/CR-5512, Volume 3, Table 6.30 (SNL 1999).
As described in Section 3.1.3, equation 3-6 includes a term to account for the net effect of all
resuspension processes (B'v) on the radionuclide concentration in vegetables. For this analysis, a
lognormal distribution with a mean of 0.001 and a standard deviation of 2.2 was assumed for B'v,
as recommended by NCRP 1999, Table 5.7, for vegetables grown in a heavily vegetated rural
area.
For this analysis, the length of the vegetable growing season (tv) was assumed to be 60 days, as
recommended in the CAP88 Users Manual (Trinity 2007).
The vegetable productivity or yield (7V) was obtained from the U.S. Department of Agriculture's
(USDA's) 2010 vegetable summary (USDA 2011). In the 2010 vegetable summary, the USDA
presents the area harvested and production values for 21 different vegetables grown in 2008,
2009, and 2010. The vegetables included in the USDA report were artichokes, asparagus, snap
beans, broccoli, cabbage, carrots, cauliflower, celery, sweet corn, cucumbers, garlic, head
lettuce, leaf lettuce, romaine lettuce, onions, bell peppers, chili peppers, pumpkins, spinach,
squash, and tomatoes. The data from the USDA report were converted into a cumulative
distribution of the vegetable yield, as shown in Figure 3-7. This is the vegetable yield (7V)
distribution that was used for this analysis. For comparison, Figure 3-7 also presents the
cumulative distribution of vegetable yield for non-leafy vegetables obtained from NUREG/CR-
6697, Appendix C, Section 6.5 (ANL 2000), and shows that the NUREG/CR-6697 yield is
substantially lower than the USDA's yield. For example, at the 50th percentile, the NUREG/CR-
6697 yield is 1.75 kg/m2, while the USDA's yield is 3.61 kg/m2. Of course, the NUREG/CR-
6697 yield is for non-leafy vegetables, which tend to have lower yields than leafy vegetables
included in the USDA's yield distribution. However, since the yield is being used in equation 3-6
to calculate the vegetable radionuclide concentration due to leaf deposition, it is believed to be
correct to include leafy vegetables in the yield cumulative distribution. Finally, Figure 3-7 shows
that the head lettuce (as a representative of leafy vegetables) has a yield of about 4.1 kg/m2.
3-21
-------
2
0.
E
u
Vegetable Yield (kg/in2)
Figure 3-7: Distribution of Vegetable Yields
The age-dependent vegetable consumption rates (Ingyeg^) were obtained by multiplying the
weight normalized vegetable consumption rates provided in EFH, Table 9-4 (EPA 201 Ib) and
shown in Table 3-9, by the age-specific body weight, as shown in Table 3-6.
Table 3-9: Age-Dependent Weight
Normalized Vegetable Consumption Rates
Percentile
1%
5%
10%
25%
50%
75%
90%
95%
99%
100%
Consumption Rate (g/kg-day)
Infant
-------
analysis and compares them to the Table 3-9 EFH distributions. As shown, there is very good
agreement between the EFH distributions and the distributions used in this analysis.
o
*
-Infant. EFH —*---Infant. CB
-Child. EFH —a—Child. CB
-Teen. EFH «-*—Teen. CB
Adult EFH Adult. CB
10 15 20
VegetableConsumption (g/ kg-day)
Figure 3-8: Distribution of Age-Dependent Weight Normalized
Vegetable Consumption Rates
As stated above, the Table 3-6 body weight distributions and the Table 3-9 vegetable
consumption distributions were used together to calculate the vegetable consumption rates
(Ingveg,A) used in this analysis.
3.2.6 Ingestion of Milk
As shown in Section 3.1.4, in order to calculate the PDCFMk,A and PRCFMk,A-, it is necessary to
know:
The element-specific cow's intake to milk transfer coefficient
The consumption rate of water by the cow (IngMk,w)
The consumption rate of fodder by the cow (Inguk.Fod)
The fraction of deposited activity retained on fodder (r/)
The length of the fodder growing season (tj)
The fodder productivity or yield (I/)
The element-specific fodder uptake factor from soil (Bf)
The net effect of all resuspension processes on fodder
The ingestion rate of soil by the cow (IngMk,Soii)
The age-dependent milk ingestion rates (IngMk,A)
3-23
-------
The following terms used to calculate the dose and risk due to milk ingestion have already been
defined in Section 3.2.4:
• The irrigation rate (Irr)
• The removal rate from soil (Aw)
• The duration period for irrigation (//,)
• The effective "surface density" for soil (P)
• The removal rate constant for physical removal by weathering (Aw)
The element-specific cow's intake-to-milk transfer coefficient (CFMk) and the element-specific
fodder uptake factor from soil (Z?/) are shown in Table 3-8, and were taken from NCRP 1999,
Appendix D.
As discussed in Section 3.2.5, NCRP Report No. 129 (NCRP 1999) includes a term to account
for the net effect of all resuspension processes on the radionuclide concentration of fodder (£/).
For this analysis, a lognormal distribution with a mean of 0.05 and a standard deviation of 1.4
was assumed for the (B'f), as recommended by NCRP 1999, Table 5.7, for fodder grown in a
heavily vegetated rural area.
The consumption rates of water (IngMk,w) and fodder (IngMk,Fod) by cows were based on
information provided by the National Academy of Sciences' National Research Council (NAS
2001) and New Mexico State University (Looper and Waldner 2002). Fodder consumption rates
are a function of the weight of the animal and how much milk the cow produces. Looper and
Waldner (2002) present dry matter intakes for a typical sized dairy cow for various milk
production rates, which range from 42 to 60 Ibs/day. Thus, for this analysis, the consumption rate
of fodder by dairy cows (IngMk.Fod) was specified as a uniform distribution, with minimum and
maximum values of 19.1 and 27.2 kg/day, respectively.
NAS 2001 also indicates that cattle water consumption rates are dependent on the dry matter
(i.e., fodder) intake, temperature, milk production, and the amount of dietary sodium. Looper and
Waldner (2002) indicate a range of dairy cow water consumption rates between 18.4 and 35.6
gallons/day. Thus, for this analysis, the consumption rate of water by cattle (IngMt,w) was
specified as a uniform distribution, with minimum and maximum values of 69.7 and
134.8 L/day.
The ingestion rate of soil by the cow (IngMk,Soii) was based on NCRP 1999, Section 5.2.3, which
states, "estimates in the literature indicate that on average an animal on pasture all year long will
ingest an amount of soil equivalent to about six percent of its total dry matter intake."
The fraction of deposited activity retained on fodder (r/) was assumed to have a triangular
distribution, with minimum, maximum, and most likely values of 0.06, 0.95, and 0.67,
respectively, as recommended in NUREG/CR-6697, Attachment C, Section 6.7.
The length of the fodder growing season ((/) was assumed to be 30 days, as recommended in
Section 10 of the CAP88 Users Manual (Trinity 2007).
3-24
-------
The fodder productivity or yield (Yj) was obtained from the U.S. Department of Agriculture's
2007 Census of Agriculture (USDA 2009). In the census, the USDA presents area and quantity
harvested in each of the 50 states for 6 types of animal feed: forage (36.0%), hay (32.0%),
alfalfa-hay (15.5%), other-tame-hay (12.5%), wild-hay (2.1%), and small-grain-hay (1.8%).
From these data, it was possible to calculate the yields and to construct a cumulative distribution
of the yield, as shown in Figure 3-9. This is the fodder yield (I/) distribution that was used for
this analysis. For comparison, CAP88 Users Manual (Trinity 2007, Section 10) and Regulatory
Guide 1.109 (NRC 1977, Table E-7) recommend an agricultural productivity by unit area for the
grass-cow-milk-man pathway of 0.28 and 0.7 kg/m2, respectively. These are equivalent to the 1.6
and 76.4 percentiles in Figure 3-9.
.a 60%
2
a.
at
E
90thPercentile
CAP88
Fodder (Dry) Yield (kg/m2)
Figure 3-9: Distribution of Fodder (Dry) Yield
The age-dependent milk consumption rates (IngMkA) were obtained by multiplying the weight
normalized milk consumption rates from EFH, Table 11-3 (EPA 201 Ib), and shown in Table
3-10, by the age-specific body weight, as shown in Table 3-6.
3-25
-------
Table 3-10: Age-Dependent Weight
Normalized Milk Consumption Rates
Percentile
1%
5%
10%
25%
50%
75%
90%
95%
99%
100%
Consumption Rate (g/kg-day)
Infant
<1 yr
0
0
0
1.2
6.4
11.5
19.6
43.2
83.1
163.9
Child
3-5 yrs
0.9
4.5
8.3
13.6
20.7
32
41.9
51.1
68.2
154.5
Teen
13-19 yrs
0.1
0.4
0.6
1.6
4
7.6
12.3
16.4
24.9
45.0
Adult
20-49 yrs
0
0.2
0.4
1
2.4
4.7
8.1
10.3
17.1
52.7
Source: EPA 20lib, Table 11-3
In order to demonstrate that the milk consumption rate data from the EFH was entered into and
used correctly by Crystal Ball, Figure 3-10 shows the distributions of the four age-dependent
milk consumption rates calculated by the Crystal Ball Monte Carlo simulation for this analysis
and compares them to the Table 3-10 EFH distributions. As shown, there is very good agreement
between the EFH distributions and the distributions used here.
-Infant EFH —•!—-Infant, CB
-Child, EFH —Q—Child, CB
-Teen, EFH —i—Teen. CB
Adult, EFH Adult, CB
25 50
Milk Consumption (g / kg-day)
Figure 3-10: Distribution of Age-Dependent Weight Normalized
Milk Consumption Rates
3-26
-------
As stated above, the Table 3-10 and Figure 3-10 milk consumption rates are presented as a
function of the body weight of the individual consuming the milk. Therefore, Table 3-6 presents
the age-dependent body weight distributions that were obtained from the EFH (EPA 201 Ib),
Table 8-3. The Table 3-6 body weight distributions and the Table 3-10 milk consumption
distributions were used together to calculate the milk consumption rates used in this analysis.
3.2.7 Ingestion of Meat
As shown in Section 3 . 1 .5, in order to calculate the PDCFMtA and PRCFMtA it is necessary to
know:
• The cattle intake-to-meat transfer coefficient
• The consumption rate of water by cattle (IngMt.w)
• The consumption rate of fodder by cattle (IngMt.Fod)
• The ingestion rate of soil by cattle (IngMt.Soil)
• Age-dependent meat ingestion rates (IngMt,A)
The following terms used to calculate the dose and risk due to meat ingestion have already been
defined in Section 3.2.6:
• The fraction of deposited activity retained on fodder (r/)
• The length of the fodder growing season (tj)
• The fodder productivity or yield (Y/)
• The element-specific fodder uptake factor from soil (Bf)
• The net effect of all resuspension processes on fodder ( B'f )
The following terms used to calculate the dose and risk due to meat ingestion have already been
defined in Section 3.2.4:
• The irrigation rate (Irr)
• The removal rate from soil (Aw)
• The duration period for irrigation (^)
• The effective "surface density" for soil (P)
• The removal rate constant for physical removal by weathering (Aw)
The element-specific cattle intake-to-meat transfer coefficient (CFj^t) is shown in Table 3-8, and
was taken from NCRP 1999, Appendix D.
The consumption rates of water (IngMt.w) and fodder (IngMt.Fod) by cattle were based on
information provided by the National Research Council (NAS 2000) and the Oklahoma State
University (Lalman 2004). Fodder consumption rates are a function of the weight of the animal,
the nutrition content of the feed, and how much the animal is growing. Table 5 of Lalman (2004)
presents dry matter intakes for various sized animals and various growth rates, which range from
13.7 to 27.4 Ibs/day. Thus, for this analysis, the consumption rate of fodder by cattle
3-27
-------
was specified as a uniform distribution, with minimum and maximum values of 6.2 and 12.4
kg/day, respectively.
NAS 2000 indicates that cattle water consumption rates are dependent on the dry matter (i.e.,
fodder) intake, temperature, precipitation, and the amount of dietary salt. Table 6-1 of NAS 2000
indicates a range of "finishing" cattle water consumption rates between 22.7 and 78.0 L/day.
Thus, for this analysis, the consumption rate of water by cattle (IngMt,w) was specified as a
uniform distribution, with minimum and maximum values of 22.7 and 78.0 L/day.
The ingestion rate of soil by cattle (IngMt,s0n) was based on NCRP 1999, Section 5.2.3, which
states, "estimates in the literature indicate that on average an animal on pasture all year long will
ingest an amount of soil equivalent to about six percent of its total dry matter intake."
The age-dependent meat consumption rates (IngMt,A) were obtained by multiplying the weight
normalized meat consumption rates from EFH, Table 11-3 (EPA 201 Ib), and shown in Table
3-11, by the age-specific body weight, as shown in Table 3-6.
Table 3-11: Age-Dependent Weight Normalized Meat
Consumption Rates
Percentile
1%
5%
10%
25%
50%
75%
90%
95%
99%
100%
Consumption Rate (g/kg-day)
Infant
<1 yr
0
0
0
0
0
1.7
3.6
5.4
9.3
18.7
Child
3-5 yrs
0
0.7
1.4
2.1
3.3
5
7.6
8.5
12.4
19.5
Teen
13-19 yrs
0
0.3
0.6
1
1.7
2.7
3.8
4.7
6.8
13.5
Adult
20-49 yrs
0
0.3
0.5
1
1.6
2.4
3.4
4.1
5.7
12
Source: EPA20lib, Table 11-3
In order to demonstrate that the meat consumption rate data from the EFH was entered into and
used correctly by Crystal Ball, Figure 3-11 shows the distributions of the four age-dependent
meat consumption rates calculated by the Crystal Ball Monte Carlo simulation for this analysis
and compares them to the Table 3-11 EFH distributions. As shown, there is very good agreement
between the EFH distributions and the distributions used in this analysis.
3-28
-------
Infant, EFH —Infant. CB
Child EFH —-s— child CB
Teen. EFH — *— Teen. CB
Adult. EFH Adult. CB
Meat Consumption (g / kg-day)
Figure 3-11: Distribution of Age-Dependent Weight Normalized
Meat Consumption Rates
As stated above, the Table 3-11 and Figure 3-11 meat consumption rates are presented as a
function of the body weight of the individual consuming the meat. Therefore, Table 3-6 presents
the age-dependent body weight distributions that were obtained from the EFH (EPA 201 Ib),
Table 8-3. The Table 3-6 body weight distributions and the Table 3-11 meat consumption
distributions were used together to calculate the meat consumption rates used in this analysis.
3.3 Pathway Dose and Risk Conversion Factors
As described in Section 3.1.7, the Excel add-in Crystal Ball was used to solve for the PDCFs and
PRCFs using Monte Carlo simulations. To solve for the PDCFs and PRCFs, the Monte Carlo
simulations were performed 1,000 times, and the resultant mean, median, and 90th percentile
PDCFs were calculated as shown in Table 3-12. The resultant PRCFs are shown in Table 3-13.
Table 3-12: Calculated Total Pathway Dose Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
Mean PDCF (mrem/yr / pCi/m3)
6.28E-03
1.30E-02
6.02E-04
1.61E-04
1.55E-04
7.33E-03
1.98E-02
3.62E-04
1.53E-04
1.47E-04
1.10E-02
2.62E-02
2.89E-04
1.33E-04
1.40E-04
3.89E-02
7.06E-02
4.27E-03
4.19E-04
4.24E-04
3-29
-------
Table 3-12: Calculated Total Pathway Dose Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
Median PDCF (mrem/yr / pCi/m3)
5.16E-03
1.01E-02
5.30E-04
1.40E-04
1.35E-04
5.17E-03
1.39E-02
2.76E-04
1.15E-04
1.10E-04
6.91E-03
1.65E-02
2.54E-04
1.01E-04
1.06E-04
3.77E-02
5.76E-02
4.37E-03
4.11E-04
4.16E-04
90th Percentile PDCF (mrem/yr / pCi/m3)
1.06E-02
2.23E-02
1.09E-03
2.77E-04
2.68E-04
1.31E-02
3.55E-02
7.43E-04
2.75E-04
2.64E-04
1.40E-02
4.28E-02
5.09E-04
1.93E-04
2.02E-04
6.31E-02
1.14E-01
7.28E-03
6.97E-04
7.06E-04
Table 3-13: Calculated Total Pathway Risk Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
Adult
Teen
Child
Infant
Mean PRCF (LCF/yr / pCi/m3)
1.12E-09
2.25E-09
3.07E-11
2.21E-11
2.28E-11
1.62E-09
3.83E-09
4.52E-11
4.22E-11
4.91E-11
2.90E-09
7.25E-09
4.69E-11
6.92E-11
9.00E-11
4.33E-09
7.60E-09
1.31E-10
1.13E-10
1.54E-10
Median PRCF (LCF/yr / pCi/m3)
9.24E-10
1.73E-09
2.69E-11
1.91E-11
1.98E-11
1.14E-09
2.69E-09
3.44E-11
3.16E-11
3.67E-11
1.82E-09
4.56E-09
4.13E-11
5.23E-11
6.80E-11
4.20E-09
6.23E-09
1.34E-10
1.11E-10
1.51E-10
90th Percentile PRCF (LCF/yr / pCi/m3)
1.89E-09
3.86E-09
5.60E-11
3.79E-11
3.92E-11
2.88E-09
6.88E-09
9.27E-11
7.58E-11
8.80E-11
3.69E-09
1.19E-08
8.27E-11
l.OOE-10
1.30E-10
7.00E-09
1.23E-08
2.22E-10
1.89E-10
2.56E-10
LCF = latent cancer fatality
Comparing Table 3-12 to Table 3-13 shows that the dose-to-risk relationship varies by a little
over an order of magnitude, from about 3xlO~8to 6xlO~7 latent cancer fatality per millirem,
depending on the radionuclide and the age group. This is consistent with the dose-to-risk
relationship for the FGR 13 dose and risk coefficients shown in Table 3-2. Furthermore, the
Interagency Steering Committee on Radiation Standards (ISCORS 2002) recommends a nominal
3-30
-------
cancer mortality dose-to-risk factor of 6x 10~7 per millirem, which is at the high end of the range,
based on the Table 3-12 PDCFs and the Table 3-13 PRCFs. However, ISCORS 2002 points out
that the nominal conversion factor varies from radionuclide to radionuclide and may
overestimate the risk by about an order of magnitude for some of bone-seeking radionuclides.
Since radium and uranium are bone-seekers, the Table 3-12 PDCFs and Table 3-13 PRCFs are
consistent with the findings in ISCORS 2002.
"Typical" contributions to the adult, teen, child, and infant total PDCFs and PRCFs from each
exposure pathway are presented in Table 3-14, Table 3-16, Table 3-17, and Table 3-18,
respectively. The contributions shown on these tables were calculated deterministically using the
mean or 50th percentile parameter values, so for any 1 of the 1,000 Monte Carlo trials, any 1 of
the pathways would be expected to contribute more or less to the dose/risk than is shown.
Nonetheless, the following tables provide a good indication as to which pathways are the major
contributors and which are only minor contributors.
Table 3-14: Typical Pathway Contributions to the Adult PDCF and PRCF
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210
78.3%
0.1%
13.5%
2.3%
0.3%
1.6%
0.3%
0.1%
0.3%
2.2%
0.5%
0.5%
Ra-226+P
51.2%
0.2%
8.8%
26.5%
0.2%
1.3%
1.1%
0.3%
0.7%
4.8%
4.1%
1.0%
Th-230
82.8%
0.3%
14.3%
2.1%
0.0%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.2%
U-234
76.2%
0.3%
13.1%
2.9%
0.3%
1.5%
0.3%
0.3%
0.4%
2.8%
0.6%
1.2%
U-238
76.2%
0.3%
13.1%
2.9%
0.3%
1.5%
0.3%
0.3%
0.4%
2.8%
0.6%
1.2%
For the adult PDCFs and PRCFs, Table 3-14 indicates that the drinking water pathway is the
major contributor for all radionuclides. Ingestion of vegetables is the second most important
contributor for all of the radionuclides. For Ra-226, the root uptake portion of the vegetable
ingestion pathway is significant, due to radium's high concentration soil uptake factor for
vegetables (Bv\ as indicated in Table 3-8. For the other radionuclides, the leaf deposition portion
of the ingestion pathways is always the higher contributor. For all radionuclides, the milk and
meat pathways are only minor contributors to the adult total PDCFs and PRCFs. Finally, the
inadvertent ingestion of soil is a very minor contributor to the adult total PDCFs and PRCFs for
all radionuclides.
The percentage contributions shown in Table 3-14 are only typical, and will vary for each of the
1000 realizations that was calculated. Table 3-15 shows how this variation in each pathway's
contribution to the total adult U-238 PDCF and PRCF. For each pathway, Table 3-15 shows the
typical contribution to the total PDCF/PRCF from Table 3-14, as well as the contributions at the
3-31
-------
10th, 50th, and 90th percentiles of the PDCF/PRCF, and the range of the contribution over all 1000
realizations.
Table 3-15: U-238+P Pathway Contributions to the Adult PDCF and PRCF
Pathway — Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Typical
(Table 3-14)
76.2%
0.3%
13.1%
2.9%
0.3%
1.5%
0.3%
0.3%
0.4%
2.8%
0.6%
1.2%
Percentile
10th
86.0%
0.5%
5.3%
0.4%
0.1%
0.4%
0.2%
0.2%
0.6%
3.7%
1.9%
0.6%
50th
95.0%
0.2%
1.3%
2.1%
0.1%
0.3%
0.1%
0.1%
0.0%
0.2%
0.1%
0.4%
90th
60.0%
0.4%
10.8%
5.7%
0.4%
0.8%
1.8%
1.4%
1.6%
2.2%
5.0%
9.9%
Range
Minimum
0.3%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Maximum
99.1%
5.5%
79.2%
68.9%
8.8%
52.8%
14.9%
7.1%
10.3%
67.1%
23.8%
49.6%
The range values shown in Table 3-15 may be misleading, since they may represent a single
Monte Carlo realization. Alternatively, Figure 3-12 shows the cumulative distribution of each
exposure pathway's contribution to the total adult U-238 PDCF/PRCF. For example, Figure 3-12
shows that for 50% of the realizations the drinking water exposure pathway contributes less than
77% and for 50% of the realizations it contributes more than 77% to the total PDCF/PRCF.
Similarly, Figure 3-12 shows that for 80% of the realizations the leaf deposition component of
the vegetable ingestion pathway contributes 21% or less to the total PDCF/PRCF. Finally, Figure
3-12 shows that for only 19% of the realizations did the drinking water pathway contribute less
than 50% to the total PDCF/PRCF.
3-32
-------
Drinking Water
Vegetables- Leaf Deposition
Vegetables- Root Uptake
Milk- Leaf Deposition
Milk- Soil Ingestion
Meat - Leaf Deposition
Milk- Root Uptake
Meat- Drinking Water
Meat- Root Uptake
Meat - Soil Ingestion
Milk- Drinking Water
Soil Ingestion
Contribution to Total PDCF/PRCF
Figure 3-12: Cumulative Distributions of Pathway Contribution to the
Total Adult U-238+P PCDF/PRCF
Although Table 3-15 and Figure 3-12 are specific for the U-238+P adult PDCF/PRCF it is
reasonable to assume that the pathway contributions for the other radionuclides and age groups
will behave similarly.
Table 3-16: Typical Pathway Contributions to the Teen PDCF and PRCF
Pathway - Teen
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210
71.5%
0.3%
14.7%
2.5%
0.4%
2.2%
0.5%
0.2%
0.7%
4.9%
1.1%
1.2%
Ra-226+P
41.4%
0.3%
8.5%
25.6%
0.3%
1.6%
1.4%
0.3%
1.3%
9.4%
8.0%
1.9%
Th-230
79.6%
0.6%
16.4%
2.4%
0.1%
0.3%
0.1%
0.1%
0.0%
0.1%
0.0%
0.4%
U-234
68.1%
0.5%
14.0%
3.1%
0.4%
2.1%
0.4%
0.4%
0.9%
6.2%
1.3%
2.6%
U-238
68.1%
0.5%
14.0%
3.1%
0.4%
2.1%
0.4%
0.4%
0.9%
6.2%
1.3%
2.6%
The pathway contributions to the teen total PDCFs and PRCFs are similar to the adult pathway
contributions. Specifically, drinking water is the major contributor for all radionuclides.
Ingestion of vegetables is the second most important contributor for all radionuclides, and
3-33
-------
inadvertent ingestion of soil is a very minor contributor for all pathways, as are the cattle/cow
drinking water and soil ingestion for the meat and milk pathways.
The milk pathway (and to a lesser extent, the meat pathway) starts to become more of a
contributor to the teen PDCFs/PRCFs than it was for the adult. For example, for Ra-226, the
milk pathway contributes about 20% to the teen total PDCF/PRCF, but only about 11% to the
adult total PDCF/PRCF. Similar relative increases are observed for the other radionuclides.
Table 3-17: Typical Pathway Contributions to the Child PDCF and PRCF
Pathway - Child
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210
59.9%
0.7%
15.6%
2.6%
0.3%
2.0%
0.4%
0.2%
1.6%
11.4%
2.5%
2.9%
Ra-226+P
28.1%
0.7%
7.3%
22.0%
0.2%
1.1%
1.0%
0.2%
2.5%
17.9%
15.3%
3.7%
Th-230
74.1%
1.9%
19.3%
2.9%
0.1%
0.3%
0.1%
0.1%
0.0%
0.2%
0.0%
1.0%
U-234
54.2%
1.4%
14.1%
3.1%
0.3%
1.8%
0.4%
0.4%
1.9%
13.8%
2.9%
5.7%
U-238
54.2%
1.4%
14.1%
3.1%
0.3%
1.8%
0.4%
0.4%
1.9%
13.8%
2.9%
5.7%
The contributions to the child total PDCFs and PRCFs from the milk pathway illustrate the
growing importance of this pathway. For Ra-226, it is the dominant pathway, exceeding both
ingestion of drinking water and vegetables. For the other radionuclides (except Th-230), the
drinking water pathway remains the most significant pathway, but the milk pathway is next in
importance. For Th-230, the milk pathway continues to be a minor contributor to the child total
PDCF/PRCF due to its very low milk uptake (CFmiik) and pasture uptake (Bp) factors, as shown
in Table 3-8.
Table 3-18: Typical Pathway Contributions to the Infant PDCF and PRCF
Pathway - Infant
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210
88.6%
0.4%
7.6%
1.3%
0.0%
0.0%
0.0%
0.0%
Ra-226+P
58.9%
1.2%
6.1%
18.2%
0.0%
0.0%
0.0%
0.0%
Th-230
87.6%
1.8%
9.0%
1.3%
0.0%
0.0%
0.0%
0.0%
U-234
81.2%
1.6%
8.4%
1.9%
0.0%
0.0%
0.0%
0.0%
U-238
81.2%
1.6%
8.4%
1.9%
0.0%
0.0%
0.0%
0.0%
3-34
-------
Table 3-18: Typical Pathway Contributions to the Infant PDCF and PRCF
Pathway - Infant
Ingestionof Milk
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Pb-210
0.2%
1.3%
0.3%
0.3%
Ra-226+P
1.0%
7.1%
6.1%
1.5%
Th-230
0.0%
0.1%
0.0%
0.2%
U-234
0.5%
3.9%
0.8%
1.6%
U-238
0.5%
3.9%
0.8%
1.6%
For the infant total PDCFs and PRCFs, Table 3-18 indicates that the ingestion of drinking water
is the major contributor for all five radionuclides; the ingestion of vegetables is the second most
important contributor after drinking water. The inadvertent soil ingestion and meat pathways
continue to be minor contributors to total PDCFs and PRCFs.
3.4 Native American Exposures
For the five ingestion pathways, the equations described in Section 3.1 for the PDCF and PRCF
calculations also apply to the Native American; the only difference is the parameter values for
the various individual consumption rates (i.e., IngwatA ^nSsoii,A-, IngvegA IngMk,A, and/«gM^). A
detailed literature search was unable to locate usage rates that were specific to Native Americans
living in Texas, Colorado, and/or Wyoming (the locations most likely to support ISL facilities).
Thus, the analysis used the Native American usage rates, which were obtained from a number of
sources focusing on tribes in the Pacific Northwest, including DOE 1997, WSDOH 2003, Harris
and Harper 2004, Rittmann 2004, PNNL 2006, Harper et al. 2007, and Harper and Ranco 2009.
3.4.1 Ingestion of Drinking Water
According to Harper and Ranco (2009, Section 8.3.1), the Native American drinking water
consumption rate depends on where the individual lives within the United States. For those
Native Americans that live in cool and wet climates (e.g., in Maine), the conventional water
consumption rate of 2 L/day (EPA 1991) should be used. For Native Americans living in areas
with a hot, arid climate (e.g., the Columbia basin), a higher water consumption rate of 3 L/day
was recommended. For this analysis, the higher water consumption rate of 3 L/day was used.
Figure 3-13 compares this Native American water consumption rate to the water consumption
distribution given in Table 3-5 and distribution from the EFH, Table 3-11 (EPA 201 Ib).
3-35
-------
o
<
1.000 1.SOO
AdultWaterComsumption(L/yr)
Figure 3-13: Native American Drinking Water Rate Within the
Exposure Factors Handbook Distribution
As Figure 3-13 demonstrates, the annual Native American drinking water consumption assumed
for this analysis is at about the 85th percentile of the EFH distribution.
3.4.2 Ingestion of Vegetables
DOE (1997, Table 5.7) gives the Native American subsistence resident consumption rate of
fruits and vegetables as 660 grams per day, with a range of 200 to 800. Rittmann (2004,
Table A4) indicates that the breakdown of the 660 g/day is 6.6% leafy vegetables, 30.2% grains,
31.4% fruit, and 31.8% other vegetables. For this analysis, it was assumed that the entire intake
of 660 g/day was of vegetables irrigated with contaminated well water. Figure 3-14 compares
this Native American vegetable consumption rate to the vegetable consumption distribution used
to calculate the PDCFs and PRCFs, which (as explained in Section 3.2.5) is a combination of the
body weight distribution (Table 3-6) and vegetable consumption distributions (Table 3-9) from
the EFH (EPA 20lib).
3-36
-------
£1
£
o.
0>
I
D
O
200 300 400
AdultVegetableConsumption (kg/yr)
Figure 3-14: Native American Vegetable Consumption Within the
Exposure Factors Handbook Distribution
As Figure 3-14 demonstrates, the annual Native American vegetable consumption assumed in
this analysis is at about the 95th percentile of the EFH distribution.
Terry (2011) indicated that gardening and/or full-scale production of fruits and vegetables on the
Navajo Reservation is virtually nonexistent. He also indicated that fruit and vegetable
consumption (even from outside locales) is probably also lower among residents of the Navajo
Reservation than for the U.S. population as a whole. Unfortunately, Terry (2011) did not provide
any numerical estimates of the Navajo vegetable consumption that could be used in this analysis,
other than to eliminate the pathway all together.
3.4.3 Ingestion of Meat
Rittmann (2004, Table A4) provides the Native American beef consumption rate as 34 kg/yr and
another 70 kg/yr for game, broken down to 22 kg/y deer, 32 kg/y wild birds, and 16 kg/y wild
bird eggs. These values are consistent with Table 5.7 of DOE (1997), which gives the following
consumption rates: 150 g/day (55 kg/yr) animal protein, 18 g/day (7 kg/yr) upland birds,
70 g/day (26 kg/yr) waterfowl, and 45 g/day (16 kg/yr) wild bird eggs. The beef and deer
consumption rates from Rittmann (2004) are equivalent to the animal protein rates from DOE
(1997). Since game would not be expected to consume well water or fodder that was irrigated,
the beef consumption rate of 34 kg/yr was used in this analysis. However, it is recognized that
use of the meat consumption rate for northwestern Native Americans (such as those given in
Rittmann 2004 and DOE 1997) may under estimate the meat consumption of southwestern
Native Americans, due to the large amount offish consumed by northwestern Native Americans.
3-37
-------
Terry (2011) indicated that meat consumption by residents of the Navajo Reservation,
particularly lamb, mutton, and their internal organs, is much higher than for the United States
population as a whole. He noted that sheep on the Navajo Nation are not fed in feedlots or
farmyards, but rely almost exclusively on grazing, and that ingestion of (contaminated) soil by
grazing livestock is probably much higher than in other locales, owing to the semi-arid climate
and overgrazing. Unfortunately, Terry (2011) did not provide any numerical estimates of the
Navajo meat consumption that could be used in this analysis.
Although both Harper and Ranco (2009) and Harper et al. (2007) provide data on the Native
American consumption of game, neither provides any data regarding the consumption of beef by
Native Americans, so they were not used in this analysis.
Figure 3-15 compares this Native American beef consumption rate to the meat consumption
distribution used to calculate the PDCFs and PRCFs, which (as explained in Section 3.2.7) is a
combination of the body weight distributions and meat consumption distributions from the EFH
(EPA 20 lib).
f
a
Jiatlve Prob
o
20% -
— —
!
: /
/
/
/
/
/'
/
/ d
/ CD
/ •£ -
/ m .5-
/ I"
«s
/ l!
0 50 100 150 200 250 300 350 4-00 4-50
Adult Meat/BeefConsumption (kg/yr)
Figure 3-15: Native American Meat Consumption Within the
Exposure Factors Handbook Distribution
Figure 3-15 shows that the Native American beef consumption rate is only at about the
35th percentile of the meat consumption distribution, based on EFH (EPA 201 Ib) data.
3.4.4 Exposure in Sweat Lodge
The Sweat Lodge Ceremony is a traditional Native American custom. Based on tribal
descriptions, PNNL 2006 (page 3-56) stated that between 0.5 and 3 hours/day could be spent
3-38
-------
sweat bathing, with the inside of the sweat lodge kept at 60°-80° Centigrade (145°-180°
Fahrenheit). For sweat lodge exposure, three sub-pathways were evaluated; inhalation of
airborne radionuclides, external exposure due to submersion in the airborne radionuclides, and
drinking additional water to make up for water lost while in the sweat lodge. The annual
individual effective dose (ESL.NA, mrem/yr) and risk (RSL.NA, LCF/yr) to a Native American from
time spent in the sweat lodge is given by:
IngWat,NA,sLDCIng + -^-(BNADCInh+DCSub)DSLNSL
8DFSL
RSL,NA = ^w
InSwat,NA,SL RCIng,DW
rr
f \
(BNA RC Ink + RCSub )DSL
where Cw is the radionuclide concentration in the well from which the water is taken
(pCi/m3water), Ingwat.NA.SL is the water consumed to make up for water lost while inside the sweat
lodge (m3/sweat), DCing is the dose coefficient for ingestion (mrem/pCi), MSL is the moisture
content of the air inside the sweat lodge (kg/m3air), S is the density of water (kg/ m3water), DFsL is
the evaporation decontamination factor (see below), BNA is the Native American's breathing rate
while inside the sweat lodge (m3air/sec), DCinh is the dose coefficient for inhalation (mrem/pCi),
DC sub is the dose coefficient for submersion within a cloud (mrem/hr per pCi/m3), DSL is the
sweat duration (hr/sweat), NSL is the number of sweats (sweats/yr), RCing,Dw is the risk coefficient
for ingestion (LCF/pCi), RCinh is the risk coefficient for inhalation (LCF/pCi), and RCsub is the
risk coefficient for submersion within a cloud (LCF/hr per pCi/m3).
As with the ingestion dose and risk coefficients, the inhalation and submersion dose and risk
coefficients were obtained from FGR 13. The inhalation dose and risk coefficients used in this
analysis are shown in Table 3-19, while the submersion dose and risk coefficients are shown in
Table 3-20.
Table 3-19: Radionuclide-Specific Inhalation
Dose and Risk Coefficients
Nuclide
Po-210
Bi-210
Pb-210
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
Mortality Coefficient (Bq"1)
Infant
0-5
2.08E-06
6.61E-08
2.67E-06
3.67E-09
4.68E-09
4.42E-06
9.24E-06
4.36E-06
4.37E-06
4.68E-09
3.77E-06
Child
5-15
1.13E-06
3.56E-08
1.23E-06
2.37E-09
3.00E-09
2.17E-06
4.57E-06
2.15E-06
2.14E-06
2.26E-09
1.82E-06
Teen
15-25
4.56E-07
1.41E-08
3.51E-07
1.09E-09
1.48E-09
7.59E-07
1.58E-06
1.03E-06
7.49E-07
8.88E-10
6.27E-07
Adult
25-70
2.20E-07
6.83E-09
2.64E-07
4.85E-10
6.12E-10
4.53E-07
9.45E-07
5.06E-07
4.47E-07
3.76E-10
3.81E-07
3-39
-------
Table 3-19: Radionuclide-Specific Inhalation
Dose and Risk Coefficients
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Effective Dose Coefficient (Sv/Bq)
Infant
100 days
1.78E-05
5.60E-07
1.83E-05
3.67E-05
9.24E-08
6.92E-08
3.35E-05
7.03E-05
2.07E-04
3.31E-05
4.08E-08
2.85E-05
2.85E-05
Child
5 yrs
8.62E-06
2.72E-07
1.15E-05
2.04E-05
3.26E-08
2.81E-08
1.89E-05
3.93E-05
.41E-04
.87E-05
.69E-08
.61E-05
.61E-05
Teen
15 yrs
5.12E-06
1.59E-07
5.83E-06
1.11E-05
1.84E-08
1.53E-08
1.04E-05
2.15E-05
l.OOE-04
1.03E-05
9.07E-09
8.69E-06
8.70E-06
Adult
20 yrs
4.27E-06
1.33E-07
5.61E-06
l.OOE-05
1.54E-08
1.47E-08
9.51E-06
1.95E-05
1.02E-04
9.40E-06
7.69E-09
8.04E-06
8.05E-06
Source: FOR 13 CD Supplement (EPA 2002)
Table 3-20: Radionuclide-Specific Submersion
Dose and Risk Coefficients
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Po-214
Bi-214
Pb-214
Po-218
Rn-222
Ra-226
Ra-226+P
Th-230
U-234
Pa-234
Th-234
U-238
U-238+P
Mortality Coefficient (m3/Bq-s)
Infant
0-5
4.63E-20
8.08E-18
4.65E-18
1.28E-17
4.54E-19
8.65E-15
1.27E-15
4.99E-20
2.10E-18
3.29E-17
9.97E-15
1.64E-18
6.22E-19
1.04E-14
3.27E-17
2.27E-19
1.04E-14
Child
5-15
4.55E-20
7.43E-18
4.58E-18
1.21E-17
4.46E-19
8.50E-15
1.25E-15
4.91E-20
2.07E-18
3.23E-17
9.79E-15
1.61E-18
6.11E-19
1.02E-14
3.22E-17
2.23E-19
1.02E-14
Teen
15-25
3.48E-20
5.94E-18
3.49E-18
9.46E-18
3.41E-19
6.51E-15
9.58E-16
3.75E-20
1.58E-18
2.47E-17
7.5E-15
1.23E-18
4.69E-19
7.8E-15
2.46E-17
1.72E-19
7.83E-15
Adult
25-70
1.33E-20
2.63E-18
1.31E-18
3.95E-18
1.31E-19
2.50E-15
3.67E-16
1.44E-20
6.06E-19
9.46E-18
2.88E-15
4.64E-19
1.71E-19
2.99E-15
9.36E-18
5.91E-20
3.00E-15
Effective Dose
Coefficient
(Sv/s)/(Bq/m3)
3.89E-19
2.58E-16
4.51E-17
3.03E-16
3.81E-18
7.25E-14
1.10E-14
4.21E-19
1.78E-17
2.84E-16
8.41E-14
1.49E-17
6.13E-18
8.73E-14
2.95E-16
2.51E-18
8.76E-14
As with the Section 3.3 pathway conversion factors, the Native American PDCFs and PRCFs for
Ra-226 included the contribution from its short-lived progeny, which are shown in both Table
3-19 and Table 3-20 as Ra-226+P. Table 3-20 also includes external dose and risk coefficients
for Rn-222 and the two very short-lived polonium isotopes, Po-214 and Po-218. Because Rn-222
3-40
-------
is a noble gas and does not remain within the body, and because the dose coefficients for Po-214
and Po-218 are so small relative to the Ra-226 dose coefficient, these three radionuclides do not
significantly contribute to either the ingestion or inhalation exposure pathways. However, as the
Table 3-20 external dose coefficients show, they may make a contribution to the external
pathway dose and risk. Notice also, that while Pa-234 and Th-234 make very small contributions
to the internal dose and risk, as compared to their parent U-238, their contribution (particularly
Pa-234) is a significant contributor to the U-238 dose and risk for the external exposure pathway.
Table 3-21 presents the values for the other parameters used in the evaluation of Native
American exposures due to the sweat lodge pathway. The source and rationale for selection of
each of the other parameter values are discussed following Table 3-21.
Table 3-21: Other Parameters Used to
Evaluate the Sweat Lodge Pathway
Parameter
Moisture content
Number of sweats
Sweat duration
Rehydration water
Breathing rate
Decontamination factor
Value
0.026
365
1
1
1.25
10
Units
kg/m3
sweats/yr"1
hr/sweat
L/sweat
m3/hr
—
DOE 1997 (Table 5.7), Harris and Harper 2004 (Table 3), Rittmann 2004 (Table A17), and
Harper et al. 2007 (Table B.2) all agree that the number of sweats and the sweat duration should
be 365 sweats/yr and 1 hr/sweat, respectively.
The WSDOH 2003 (Section 3.2), Harris and Harper 2004 (Table 3), Harper et al. 2007
(Section 3.6), and Harper and Ranco 2009 (Section 8.3.1) all agree that 1 liter of water is needed
to make up for water lost while inside the sweat lodge (Ingwat,NA,SL)-
Although Harper and Ranco 2009 (Appendix 1) recommends a Native American breathing rate
of 25 m3/day, DOE 1997 (Table 5.7), WSDOH 2003 (Table 3.2.2), Rittmann 2004 (Table A10),
and Harper et al. 2007 (Table B. 1) all recommend that 30 m3/day be used as the adult Native
American breathing rate. Thus, for this analysis, a Native American breathing rate of 30 m3/day
was assumed.
Many previous sweat lodge exposure analyses (e.g., DOE 1997, WSDOH 2003, ATSDR 2010)
have based their values for the last two parameters, the sweat lodge air moisture content (MSL)
and the sweat lodge evaporation decontamination factor (DFsi), on a methodology developed by
Rodney S. Skeen, PhD, Confederated Tribes of the Umatilla Indian Reservation (CTUIR),
Department of Science and Engineering, and presented in Harper et al. 2007 and elsewhere.
Recently, the Skeen methodology has come under criticism. For example, in the River Corridor
Baseline Risk Assessment (DOE 2010), the Hanford risk assessors expressed concern with using
the Skeen methodology:
3-41
-------
There is no physical basis provided in Harris and Harper (...) to correlate
respirable aerosol concentrations with saturated water vapor, and exposure
concentrations ofnonvolatiles in sweat lodge air calculated in this manner are
considered physically implausible, (page 3-71)
The Harris and Harper (...) equation for calculating air-phase EPCs [exposure
point concentrations] for nonvolatile analytes (Equation 3-2) calculates the
concentration of a nonvolatile COPC [contaminant of potential concern] in air as
a function of the concentration of water vapor produced by the volatilization of
water poured over hot rocks in a sweat lodge. Because nonvolatile contaminants
have no vapor pressure, Equation 3-2 does not have a common physical basis
with volatile chemicals, (page 6-77)
Furthermore, on page O-60, Rittmann (1999) said the following about the Skeen methodology as
utilized by the Columbia River Comprehensive Impact Assessment (DOE 1997):
..., the sweat lodge temperature is 60° C (140° F) with a relative humidity of
100%, which means the air in the sauna is 20% water vapor. The sudden 20%
reduction in oxygen concentration would certainly result in labored breathing,
possibly fainting. The exposure to high temperature at saturated conditions for
one hour would lead to skin burns over most of the body.
The Agency for Toxic Substances and Disease Registry (ATSDR) prepared a public health
assessment (PHA) for the Midnite Mine site, and included an evaluation of sweat lodge
exposures that utilized Skeen's methodology. Among the comments received on the ATSDR
Midnite Mine PHA was the following (ATSDR 2010, page E-23):
Yet another flaw in the sweat lodge models relates to the assumption that
nonvolatile chemicals in water brought into the sweat lodge will be released to
the air. During use of a sweat lodge, the PHA assumes that water is poured over
heated rocks to generate steam or water vapor via the evaporation process.
Evaporation is defined as a process whereby atoms or molecules in a liquid state
gain sufficient energy to enter the gaseous state. For evaporation to occur, the
atoms or molecules of a chemical component in a liquid solution must be heated
to a level that reaches the boiling point for that chemical. Although the heated
rocks in the sweat lodge could be hot enough to evaporate the water, they would
not be hot enough to vaporize the manganese and other metals present in the
water. The boiling points for most of the metals are extremely high. For example,
the boiling point for elemental manganese is 2,061° C (3,742° F). Chemical
components with such high boiling points will remain in the solution (via
processes of distillation and condensation) or will form a precipitate. Therefore,
the assumption that the water vapor inside the sweat lodge contains metals at the
same concentrations as in the water poured on the rocks is flawed.
WSDOH (2003, Section 5.0.2) had this to say regarding Skeen's methodology for calculating the
sweat lodge airborne concentration:
3-42
-------
The operating assumption is that 100% of the contaminants in the groundwater
(used as the source of steam for the sweat lodge) will become airborne and
remain available for inhalation. Uranium andplutonium compounds have a
higher melting point than the temperature observed in a sweat lodge and must be
entrained in the water transitioning to steam to be available for inhalation. Of
those contaminant particles in the air, it is likely that the deposition rate will be
higher than that of water vapor and would also serve to decrease the average air
concentration. In addition, it is likely that a fraction of the contaminants will fail
to become entrained in the water and become airborne, further reducing the air
concentrations from those used in the calculation. The sweat lodge calculations
are therefore considered a worst case estimate of the potential exposure to
contaminants. Until data are available on the potential air concentration in a
similar environment, the current model is considered the appropriate method for
estimating exposure.
For this analysis, the Skeen methodology was set aside and the airborne radionuclide
concentrations were calculated based on realistically conservative assumptions regarding the
sweat lodge air moisture content (MSL) and the sweat lodge evaporation decontamination factor
Sweat Lodge Air Moisture Content
The first of Skeen' s assumptions that was challenged is that the air within the sweat lodge is at
100% humidity. While this is clearly a conservative assumption, it leads to conditions inside the
sweat lodge that would make the sweat lodge uninhabitable. That is, at sweat lodge temperatures
(i.e., between 60° C and 80° C), the humidity must be kept down to nearly zero in order to
prevent scalding of the skin on contact with the air moisture. When the humidity approaches
100%, a much lower temperature of around 40° C (104° F) is necessary to prevent scalding. This
analysis utilized a humidity level that, while conservative, is consistent with a habitable sweat
lodge.
The heat index combines air temperature and relative humidity in an attempt to determine the
human-perceived equivalent temperature, i.e., how hot it feels. When the heat index is above
80° F, the National Weather Service (NWS) will issue a caution, and the NWS considers a heat
index above 130° F to be extremely dangerous, with the possibility of heatstroke highly likely.
At the above sweat lodge temperature range, the assumption that the air is saturated (i.e., 100%
humidity) would result in a heat index well above 500° F, which is unreasonable. Thus, the
humidity within a sweat lodge must be considerably less than 100%.
For different temperatures, Table 3-22 shows what the moisture content of the air would be when
the heat index is at 130° F. The maximum air moisture content (0.028 kg/m3) is achieved at the
lowest temperature (100° F). Because the Sweat Lodge Ceremony pushes the individuals beyond
their normal limits, Table 3-22 also shows what the air moisture content would be when the heat
index is at 180° F, significantly higher than the extremely dangerous level. With a heat index of
180° F, the moisture content of the air is 0.026 kg/m3 or less in the sweat lodge temperature
range (i.e., 145° to 180°F).
3-43
-------
Table 3-22: Air Moisture Content with Heat
Indices of 130° F and 180° F
Temp
(OF)
100
110
120
130
140
145
150
155
160
165
170
175
180
Heat Index: 130° F
RH
60.4
36.0
20.1
11.1
6.65
5.40
4.50
3.85
3.41
3.10
2.87
2.72
2.62
Moisture
(kg/m3)
0.028
0.022
0.016
0.011
0.0086
0.0079
0.0074
0.0071
0.0071
0.0072
0.0074
0.0078
0.0084
Heat Index: 180° F
RH
92.3
64.1
44.1
30.2
21.1
17.8
15.3
13.2
11.6
10.3
9.3
8.4
7.7
Moisture
(kg/m3)
0.042
0.039
0.035
0.031
0.027
0.026
0.025
0.024
0.024
0.024
0.024
0.024
0.025
For this analysis, the moisture content of the air inside the sweat lodge (Msi) was assumed to be
0.026 kg/m3, since that is the maximum over the 145° F to 180° F temperature range when the
heat index is 180° F, and it is near the maximum over the 100° F to 189° F when the heat index
is 130° F.
Sweat Lodge Evaporation Decontamination Factor (DFsi)
The second of Skeen's assumptions that was challenged is that the concentration of nonvolatiles
in the sweat lodge airborne moisture is the same as the nonvolatile ground water concentration.
Since evaporation is often used to remove impurities from water, this assumption is very
conservative. For example, providers of distilled water first evaporate the water and then re-
condense it, leaving behind virtually all of the impurities. Also, the NRC (1985, Section 2.19.1)
recommends that, depending on its use, an evaporator can remove 99% to 99.9% of contamina-
tion from the re-condensed water, resulting in decontamination factors of 100 to 1,000. While
pouring water onto hot stones in a sweat lodge is a far cry from a distiller or evaporator, it
nonetheless results in the evaporation of the water.
From Exposure Scenarios and Unit Dose Factors for the Hanford Immobilized Low-Activity
Tank Waste Performance Assessment (Rittmann 1999), page O-57:
Note that only a small fraction of the radioactivity in the evaporating water
becomes airborne. In Airborne Release Fractions/Rates and Respirable Fractions
for Nonreactor Nuclear Facilities, Volume 1 (DOE-HDBK-3010-94) data for the
sudden depressurization of superheated aqueous solutions is presented in Table
3-5 for various initial pressures and volumes. With a source volume of 0.35 L at a
pressure of60psig, the respirable release fraction is 0.006. A somewhat larger
3-44
-------
value ofO. 01 will be used to represent the resuspensionfrom pouring water on
hot rocks in the sweat lodge, and for the spray of a shower. (The only exception to
this is tritium. Since it is assumed to be oxidized, 100% of the tritium becomes
airborne.)
In the 1970s, the Oak Ridge National Laboratory performed a survey to evaluate the
effectiveness of evaporation as a treatment method for reducing releases of radioactive effluents
to the environment from nuclear power plants. The principal emphasis was placed upon data
concerning the ratio of the feed to condensate concentrations, or decontamination factor (DF).
For nonvolatile contaminants, Godbee (1973, page 2) found that:
An average system DF of JO3 to JO4 can be expected under routine operating
conditions for nonvolatile radioactive contaminants treated in evaporators.
Godbee (1973) noted that the evaporation DF could be reduced by entrainment, which is liquid
suspended in the vapor as fine droplets that are carried along with the rising vapor stream.
NUREG-1140 (NRC 1991), Section 2.3.1.2, recommends a release fraction be used when
evaluating nuclear fuel cycle facility accidents. Regarding the release of nonvolatile compounds
from non-flammable liquids (e.g., water), NUREG-1140 (page 77) states:
Nonvolatile compounds in nonflammable liquids are assigned a release fraction
of 0.001. Several studies have measured releases in these circumstances. In
general, release of these compounds can be expected to be small until the liquid is
dried. After drying release fractions generally remain small because the material
normally cakes on the substrate or binds into particles too large to be respirable.
It is recognized that none of the above examples exactly represents the pouring of water onto hot
stones in a sweat lodge. However, taken together, they provide convincing evidence that only a
small portion of nonvolatile contaminants are released during the evaporation of water. Although
the above discussion suggests that a value of either 100 or 1,000 could be used, for this analysis,
a sweat lodge evaporation decontamination factor (DFsi) of 10 was assumed. By using this low
value for the DFsL, one acknowledges that a significant fraction of nonvolatile contaminants
could become airborne, while still recognizing that, for water poured onto hot stones, the
behavior of nonvolatile contaminants within the water is poorly understood.
3.4.5 Native American Pathway Dose and Risk Conversion Factors
Using the above described models and parameter values, the calculated Native American PDCFs
and PRCFs are shown in Table 3-23.
3-45
-------
Table 3-23: Calculated Native American Total
Pathway Dose and Risk Conversion Factors
Nuclide
Pb-210+P
Ra-226+P
Th-230
U-234
U-238+P
PDCF
(mrem/yr / pCi/m3)
5.10E-03
2.68E-02
1.94E-03
4.14E-04
3.70E-04
PRCF
(LCF/yr/pCi/m3)
6.63E-10
4.86E-09
9.86E-11
7.08E-11
6.36E-11
Comparing the Native American PDCFs and PRCFs to the adult total PDCFs and PRCFs given
in Table 3-12 and Table 3-13, respectively, shows that the Native American PDCFs and PRCFs
are greater than the 90th percentile adult total PDCFs and PRCFs. As explained below in the
discussion of the pathway contributions to the Native American PDCFs/PRCFs, this is due to the
larger drinking water and vegetable consumption rates that were assumed for the Native
American.
Table 3-24 presents the breakdown by exposure pathway of the Native American PDCFs, while
Table 3-25 does the same for the Native American PRCFs. Because the Section 3.3 PDCFs and
PRCFs only included ingestion pathways, the relative pathway contributions to the total PDCF
are the same as to the total PRCF. However, since the Native American PDCFs/PRCFs include
the inhalation and ingestion pathways, and because the dose-to-risk relationship differs between
the inhalation and ingestion pathways, it was necessary to present the Native American pathway
contributions on separate tables.
Table 3-24: Native American Pathway Contributions to the PDCF
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Sweat Lodge
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Inhalation
Submersion
Drinking Water
Pb-210+P
55.3%
0.1%
18.1%
3.0%
0.1%
0.4%
0.1%
0.0%
0.3%
2.4%
0.5%
0.6%
0.5%
0.0%
18.4%
Ra-226+P
33.0%
0.2%
10.8%
32.4%
0.1%
0.3%
0.3%
0.1%
0.7%
4.9%
4.2%
1.0%
1.2%
0.0%
11.0%
Th-230
44.7%
0.2%
14.6%
2.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
23.1%
0.0%
14.9%
U-234
48.5%
0.3%
15.9%
3.5%
0.1%
0.4%
0.1%
0.1%
0.4%
2.9%
0.6%
1.2%
10.0%
0.0%
16.2%
U-238+P
48.8%
0.3%
16.0%
3.6%
0.1%
0.4%
0.1%
0.1%
0.4%
2.9%
0.6%
1.2%
9.6%
0.0%
16.3%
3-46
-------
Table 3-24 shows that the drinking water pathway is the major contributor to the PDCF for four
of the five radionuclides. For Ra-226+P, the vegetable consumption pathway (particularly the
root uptake sub-pathway) is the major contributor to the PDCF. This is due to radium's large
concentration soil uptake factor for vegetables (Bv), as indicated in Table 3-8. The sweat lodge
contribution (particularly the rehydrate drinking water sub-pathway) to the PDCF is significant
for all five radionuclides. For Th-230, U-234, and U-238, the sweat lodge inhalation sub-
pathway is also a contributor. As with the Section 3.3 PDCFs, the inadvertent ingest of soil,
meat, and milk pathways are small contributors to the Native American total PDCF.
Table 3-25: Native American Pathway Contributions to the PRCF
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of
Vegetables
Ingestion of Meat
Ingestion of Milk
Sweat Lodge
Leaf Deposition
Root Uptake
Cattle Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Cow Drinking
Leaf Deposition
Root Uptake
Soil Ingestion
Inhalation
Submersion
Drinking Water
Pb-210+P
53.6%
0.1%
18.7%
3.1%
0.1%
0.4%
0.1%
0.0%
0.4%
2.5%
0.6%
0.6%
1.8%
0.0%
17.9%
Ra-226+P
30.5%
0.2%
10.8%
32.3%
0.1%
0.3%
0.3%
0.1%
0.7%
4.8%
4.1%
1.0%
4.9%
0.0%
10.2%
Th-230
44.0%
0.3%
15.9%
2.4%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
22.6%
0.0%
14.7%
U-234
37.8%
0.2%
13.7%
3.0%
0.1%
0.3%
0.1%
0.1%
0.3%
2.5%
0.5%
1.0%
27.8%
0.0%
12.6%
U-238+P
38.6%
0.2%
13.9%
3.1%
0.1%
0.3%
0.1%
0.1%
0.3%
2.5%
0.5%
1.0%
26.4%
0.0%
12.9%
Table 3-25 shows similar results for the pathway contributions to the Native American total
PRCFs as the Table 3-24 PDCFs, except that the contribution from the sweat lodge inhalation
sub-pathway has more than doubled. The reason for this is that on a per-Curie basis, the
inhalation risk coefficients (Table 3-19) are a factor of 28 (Ra-226) to 612 (U-234) times larger
than the ingestion risk coefficients (Table 3-2).
3.5 Exposure Pathways Not Analyzed in Detail
This section describes scoping calculations, which were performed for three potential exposure
pathways: (1) radon in the home due to off-gas from well water, (2) exposure to the fetus while
in the womb, and (3) infant exposure from the consumption of breast milk. Because these
scoping calculations showed that they did not contribute significantly to the PDCFs, the three
exposure pathways discussed in this section were not analyzed in detail.
3-47
-------
3.5.1 External Exposure to Contaminated Ground
A screening calculation was performed to determine the impact from external exposure to soil relative to
the soil ingestion pathway. It was assumed that an individual would spend 136.4 minutes per day
outdoors at their residence (EPA 201 Ib, Table 16-22) (SNL 1999, Volume 3, Table 6.7 gives a slightly
larger value of 158.5 min/day [40.2 24-hr days/year] as the mean time spent outdoors at a residence). As
Table 3-26 shows, it was determined that relative to the soil ingestion pathway, the external exposure
pathway can be a significant contributor to the dose, primarily due to the contributions from the daughter
products (e.g., Bi-210, Bi-214, Pb-214, Th-234). However, based on Table 3-14, the external exposure
pathway could contribute only about 1.3% to the Ra-226+P overall PDCFs/PRCFs, and much less for the
other radionuclide.
Table 3-26: External Exposure to
Contaminated Ground Screening Study Results
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Soil Dose (Sv/yr)/(Bq/g)
External
1.26E-09
1.40E-07
5.07E-08
1.92E-07
2.39E-04
3.18E-05
7.46E-07
2.71E-04
2.74E-08
8.80E-09
5.45E-07
2.04E-09
5.47E-07
Ingestion
2.21E-05
2.39E-08
1.27E-05
3.48E-05
2.05E-09
2.54E-09
5.11E-06
4.00E-05
3.91E-06
9.04E-07
6.21E-08
8.13E-07
8.75E-07
3.5.2 Exposure to Indoor Radon
If the well water contains radium-226, then it will also contain its decay product radon-222 (see
Figure 3-2). Radon-222 is a noble gas, and once the water is in the home, the radon will escape
and become airborne within the home. Individuals within the home will be exposed to radon and
its short-lived daughter products (i.e., Po-218, Pb-214, Bi-214, and Po-214; see Figure 3-2) via
two exposure pathways; inhalation and submersion. The amount of radon airborne in the home
depends upon the amount of water brought into the home, how much of the radon escapes from
the well water, and the home's air turnover rate. The following equation was used to calculate
the exposure to radon that is released from well water brought into the home.
3-48
-------
PDCF,
"
wat
Radon
-^Lfh[EfBKDCInh+DCSub]
3-14
where PDCFRadon is the PDCF due to radon in a home where well water is used (mrem/yr
per pCi/m3water), Uwa is the per capita water usage rate (m3/day per person), 5/, is the number of
people living in the home (people),/0_g is the radon off-gas fraction, /I/, is the home's air turnover
rate (day"1), Vh is the volume of the home (m3),^//, is the fraction of time spent in the home, Efis
the radon daughter product equilibrium fraction, B is the breathing rate (m3/day), DCint, is the
radon daughter products dose coefficient for inhalation (mrem/pCi), K is a units conversion
factor (day/year), and DCsub is the dose coefficient for radon submersion within the home
(mrem/hr per pCi/m3). The value of each of the parameters used in equation 3-14 has been
tabulated in Table 3-27.
Table 3-27: Data Used to Evaluate the PDCF for Home Radon Exposure
Parameter
Per capita water usage rate
Number of people living in the home
Radon off-gas fraction
Home's air turnover rate
Volume of the home
Fraction f time spent in the home
(residence)
Breathing rate
Radon daughter product equilibrium
fraction
Radon daughter products dose
coefficient for inhalation
Dose coefficient for radon
submersion
Uwat
sh
Jo-z
h
vh
fk
B
Ef
DClnh
DCSub
Units
(nrVday per person)
(people)
None
(day'1)
(mj)
None
(nrVday)
None
(mrem/pCi)
(mrem/yr per pCi/m3)
Value
0.26
2.58
0.90
10.8
492
0.66
16.3
4.32
154
0.99
24.6
0.4
1.55E-04
2.07E-06
Source
EPA 2009
USCB2011
NAS 1999a, page 91
EPA 20 lib, Table 19-1
EPA 20 lib, Table 19-1
EPA 20 lib, Table 16-1
EPA 20 lib, Table 5-1
NAS 1999b, Figure B-12
FOR 13 CD Supplement,
EPA 2002
FOR 13 CD Supplement,
EPA 2002
For the parameters whose values were obtained from the EFH (EPA 201 Ib), two values are
presented in Table 3-27; the first value was used to calculate an estimate of the mean exposure,
while the second value was used to calculate an upper estimate of the exposure. The radon off-
gas fraction (f0_g) was used to calculate how much radon escapes from the water into the air. The
value for the off-gas fraction (90%) was taken from NAS 1999a, and is specifically for radon
escaping in the shower, although it has been applied to all water for this scoping calculation.
Other researchers have published a radon off-gas fraction closer to 70% (Hopke 2006).
The results of the radon exposure pathway PDCF scoping calculation are shown in Table 3-28.
Comparing the radon PDCFs from Table 3-28 to the Ra-226+P PDCFs from Table 3-12 shows
that the contributions from the radon exposure pathways are small (i.e., two to three orders of
magnitude lower).
3-49
-------
Table 3-28: PDCF for Home Radon Exposure
Exposure
Pathway
Inhalation
Submersion
Total
Radon PDCF (mrem/yr per pCi/m3)
Mean
3.10E-05
1.74E-10
3.10E-05
Upper
5.63E-04
2.09E-09
5.63E-04
3.5.3 Swimming Pool/Hot Tub Exposures
A screening calculation was performed to determine the impact from external exposure to water,
such as by swimming or a hot tub. For this screening calculation it was assumed that an
individual would spend 60 minutes per month in either a swimming pool or hot tub (EPA 201 Ib,
Table 16-57, Overall, 50* Percentile). As Table 3-29 shows, it was determined that the external
exposure pathways are not a significant contributor to the dose (relative to drinking water).
Based on these results, the water immersion exposure pathways were not analyzed in detail in
this analysis.
Table 3-29: Water Immersion
Screening Calculation
Nuclide
Po-210
Bi-210
Pb-210
Pb-210+P
Bi-214
Pb-214
Ra-226
Ra-226+P
Th-230
U-234
Th-234
U-238
U-238+P
Water Dose (Sv/yr)/(Bq/m3)
Immersion
3.90E-17
2.73E-15
5.66E-15
8.43E-15
7.17E-12
1.12E-12
3.00E-14
8.33E-12
1.70E-15
7.56E-16
3.30E-14
3.43E-16
3.33E-14
Drinking
6.54E-07
7.08E-10
3.76E-07
1.03E-06
6.05E-11
7.51E-11
1.51E-07
1.18E-06
1.16E-07
2.68E-08
1.84E-09
2.41E-08
2.59E-08
3.5.4 Exposures Due to Hydroponics and/or Aquaculture
Appendix V of the Background Information Document (BID) prepared by EPA for its 40 CFR
197 rulemaking (EPA 2001) addressed the impact of contaminated ground water on hydroponics
farming (i.e., the science of growing plants without soil). After reviewing the literature on
hydroponics, the BID concluded: given that hydroponically-grown vegetables would not be
subject to the buildup ofradionuclides in soil, it is reasonable to conclude that they would have
3-50
-------
lower radionuclide concentrations than vegetables grown in soil. (EPA 2001, page 8-46) It is
reasonable to make a similar conclusion for this analysis. Also, as shown in Table 3-13 through
3-16, with the exception of Ra-226, the root uptake component contributions to the total
PDCFs/PRCFs are small (e.g., less than 4%). For these reasons, a detailed assessment of the
hydroponics exposure pathway has not been included in this analysis.
Likewise, Appendix V of the Background Information Document (BID) prepared by EPA for its
40 CFR 197 rulemaking (EPA 2001) also addressed the impact of contaminated ground water on
aquaculture (i.e., fish farming). After reviewing the literature on hydroponics, the BID found
that: In arid areas, fish farming is usually conducted in large tanks filled with ground water that
is continually filtered and aerated. Food, in the form of commercial pelletizedfloating feed, is
introduced into the tanks daily. The extensive literature on concentration factors for
radionuclides in freshwater fish is not considered applicable to the unique conditions of
aquaculture. Uptake is limited to direct sorption of radionuclides in the water. (EPA 2001, page
8-46) For these reasons, a detailed assessment of the aquaculture exposure pathway has not been
included in this analysis.
3.5.5 Embryo and Fetus Exposure
While in the womb, an embryo/fetus receives nourishment from its mother. If the mother ingests
and/or inhales radioactivity during her pregnancy, then some of that radioactivity may be
delivered to the fetus. This phenomenon has been studied by the ICRP, and their results and
recommendations have been published in Publication 88 (ICRP 2001).
In Publication 88 new biokinetic and dosimetric models for calculating doses to
the developing embryo and fetus are developed .... The models [which are
developed] take account of transfer of radionuclides across the placenta,
distribution and retention of radionuclides in fetal tissues, growth of the fetus, and
photon irradiation from radionuclides in the placenta and maternal tissues. ...
Intake scenarios comprising single or continuous maternal intakes are taken into
account in the compilation of effective dose coefficients following ingestion or
inhalation of the radionuclides considered. (ICRP 2001)
Table 3-30 gives the ICRP Publication 88 effective dose coefficients from conception to birth (in
utero), as well as the effective dose coefficients from birth to age 70 years (post natal) due to
exposures while in the womb.
Table 3-30: Dose Coefficients to the Offspring from
Chronic Intake by the Mother
Nuclide
Pb-210
Ra-226
Time of Intake
(weeks)
-260
-52
0
-260
-52
0
Dose Coefficient (Sv/Bq)
In Utero
2.4E-08
3.0E-08
LIE-OS
1.4E-09
2.9E-09
2.9E-07
Post Natal
5.1E-08
7.0E-08
1.3E-07
5.0E-11
8.3E-11
2.8E-08
3-51
-------
Table 3-30: Dose Coefficients to the Offspring from
Chronic Intake by the Mother
Nuclide
Th-230
U-234
U-238
Time of Intake
(weeks)
-260
-52
0
-260
-52
0
-260
-52
0
Dose Coefficient (Sv/Bq)
In Utero
1.9E-10
2.2E-10
7.6E-10
7.6E-10
9.9E-10
LIE-OS
6.8E-10
8.8E-10
9.5E-09
Post Natal
7.5E-10
7.7E-10
7.8E-09
1.8E-10
2.1E-10
4.1E-09
1.6E-10
2.0E-10
3.8E-09
Source: ICRP 2001
The Table 3-30 dose coefficients are for chronic intake of radioactivity for three different time
periods: (1) for 5 years (260 weeks) up to (but not after) conception, (2) for 1 year (52 weeks)
up to (but not after) conception, and (3) during the pregnancy, starting from conception.
The first and third sets of dose coefficients were combined with the water consumption rate for
non-pregnant and pregnant women (i.e., 1,243 and 1,318 mL/day, respectively) from EFH (EPA
201 Ib), Table 3-76. This model assumes that the woman drinks water from the contaminated
well for 5 years prior to conception, and continues to drink from the well throughout her
pregnancy. The results of this scoping calculation are shown in Table 3-31. Comparing the
PDCFs from Table 3-31 to the PDCFs from Table 3-12 shows that the contribution to the total
exposure from the chronic intake of radioactivity by the mother during pregnancy is small.
Table 3-31: PDCF to the Offspring from
Chronic Intake by the Mother
Nuclide
Pb-210
Ra-226
Th-230
U-234
U-238
PDCF (mrem / pCi/m3)
In Utero
2.14E-04
3.58E-04
2.50E-06
1.95E-05
1.70E-05
Post Natal
5.82E-04
3.39E-05
1.56E-05
6.41E-06
5.88E-06
Total
7.96E-04
3.92E-04
1.81E-05
2.59E-05
2.29E-05
3.5.6 Infant Consumption of Formula/Milk
Infants receive their nourishment from formula, breast milk, or cows' milk. Furthermore,
formula is available in three forms; powder, liquid concentrate (which requires dilution), and
premixed ready-to-feed. Currently, 74.6% of American infants have been breastfed at least once,
with 35% of infants exclusively breastfed at 3 months, 44.3% receiving some breastfeeding at
6 months, and 23.8% receiving breastfeeding at 1 year (CDC 2011). Consumption of cows' milk
is not recommended for infants under the age of 1 year; however, that recommendation is often
ignored and cows' milk is introduced earlier. EFH (EPA 201 Ib), Table 3-72, shows that 5% of
infants consumed cows' milk at 6 months, increasing to 25% at 9 months and to 79% at
12 months.
3-52
-------
Except for premixed ready-to-feed formula, each of these means of feeding could result in
exposure to an infant if the well water is contaminated. For powder or liquid concentrate
formula, well water is mixed with the formula and fed directly to the infant. Section 3.1.1
presents the methodology used to calculate exposures due to water consumption, and this same
methodology was used to calculate exposure from powder or liquid concentrate formula
consumption. Table 3-5 and Figure 3-3 show the infant's water consumption distribution, which
was obtained from EFH (EPA 201 Ib), Table 3-41. On Figure 3-3, the weight normalized infant's
water consumption is significantly larger than the water consumption for the other age groups,
reflecting the fact that water is used to prepare formula, which is the bulk of an infant's
nourishment.
The methodology used to calculate exposures from the cows' milk pathway was described in
Section 3.1.4. Table 3-10 and Figure 3-10 show the infant's cows' milk consumption
distribution, which was obtained from EFH (EPA 201 Ib), Table 11-3. On Figure 3-10, the
weight normalized infant's milk consumption is significantly smaller than the child's milk
consumption; this reflects the fact that cows' milk is not recommended for infants under the age
of 1 year.
Thus, the infant's exposures from consumption of formula and cows' milk have been included in
the PDCFs that are presented in Section 3.3. However, the infant's exposure from breast milk
consumption has not been addressed. The ICRP has studied this exposure pathway, and their
results and recommendations have been documented in Publication 95 (ICRP 2004). In
Publication 95, the ICRP adapted existing biokinetic models for the female adult to include
transfer of radioactivity to milk. The fraction of the mother's radioactivity intake that reaches the
breast milk was then combined with the infant's ingestion dose coefficient to arrive at a dose
coefficient that relates the amount of radioactivity ingested by the mother to the dose received by
the breastfeeding infant (i.e., Sv to the infant/Bq intake by the mother). Table 3-32 shows the
fraction of mother's radioactivity intake that reaches the breast milk and the breastfeeding infant
dose coefficient obtained from ICRP Publication 95, as well as the risk coefficient, which was
calculated in the same manner as the dose coefficient.
Table 3-32: Dose and Risk Coefficients from Breast Milk Consumption
Nuclide
Pb-210
Ra-226
Th-230
U-234
U-238
Fraction of Mother's
Intake to Breast Milk*
2.5E-02
5.9E-03
2.4E-05
9.7E-04
9.7E-04
Dose Coefficient
(Sv/Bq)*
2.2E-07
2.8E-08
9.8E-11
3.6E-10
3.3E-10
Risk Coefficient
(LCF/Bq)
9.6E-09
2.5E-09
3.0E-13
9.7E-12
1.3E-11
*ICRP 2004
Table 3-33 presents a comparison of the doses and risks received by infants consuming cows'
milk, formula, and breast milk. For cows' milk and formula, it was assumed that the infant
consumed 778 mL/day (EPA 201 Ib, Table 3-74), and for breast milk, it was assumed that the
mother consumed 1,806 mL/day of water (EPA 201 Ib, Table 3-76). The radioactivity in the
cows' milk was calculated using the methodology from Section 3.1.4 and the parameter values
3-53
-------
from Section 3.2.6. In addition to consuming contaminated well water, the mother was assumed
to consume contaminated vegetables, cows' milk, and meat, as described in Sections 3.1 and 3.2.
Table 3-33: PDCF/PRCF for Infant Milk Consumption
Infant Milk Ingestion - Dose (mrem/yr / pCi/m3)
Cows' Milk
Formula
Breast Milk
Pb-210
1.28E-02
3.61E-02
6.59E-04
Ra-226
1.76E-01
4.10E-02
1.22E-04
Th-230
1.04E-04
4.34E-03
2.80E-07
U-234
3.52E-04
3.88E-04
1.10E-06
U-238
3.53E-04
3.93E-04
1.01E-06
Infant Milk Ingestion - Risk (LCF/yr / pCi/m3)
Cows' Milk
Formula
Breast Milk
Pb-210
1.43E-04
4.02E-04
7.35E-06
Ra-226
1.91E-03
4.44E-04
1.32E-06
Th230
3.20E-07
1.34E-05
8.61E-10
U-234
9.53E-06
1.05E-05
2.99E-08
U-238
1.28E-05
1.42E-05
3.66E-08
As Table 3-33 shows, the infant's exposure due to consuming breast milk is small compared to
the exposure due to either formula or cows' milk consumption. This is expected, since the
mother's body only allows a fraction of the radioactivity entering her body to reach her breast
milk (Table 3-32), where it can be consumed by the infant, whereas for formula consumption, all
radioactivity in the well water is consumed by the infant.
3-54
-------
4.0 DOSE AND RISK ASSESSMENT
4.1 Introduction
Chapter 3 described the methodology for calculating probabilistic pathway dose and risk
conversion factors, while Chapter 2 described the ground water modeling for excursion and
leakage scenarios. The ground water modeling was based on an initial source concentration of
1 mg/L for a generic species, and concentrations were calculated at down-gradient receptor wells
located various distances from the source assuming no retardation. In this chapter, we evaluate
expected doses and risks assuming source term strengths based on available measurements from
the literature and include the effects of radionuclide-specific retardation factors.
4.2 Selection of Lixiviant Concentrations and IQs
To be conservative, we have generally selected low-end KdS and high-end lixiviant
concentrations for calculating doses and risks.
4.2.1 Lixiviant Concentration
Table 4-1 summarizes reported ranges of concentrations of various species in ISL/ISR lixiviants
(PNNL2010).
Table 4-1: Representative Concentrations in Uranium Alkaline ISR Lixiviants
Constituent
Calcium
Magnesium
Sodium
Potassium
Carbonate
Bicarbonate
Chloride
Sulfate
Silica
pH (standard units)
Total dissolved solids
Alkalinity (as CaCO3)
Arsenic
Iron
Manganese
Molybdenum
Radium-226 (pCi/L)
Selenium
Uranium
Vanadium
Wyoming
Site(a)
138
42
365
12
NR
NR
140
229
24.6
6.7
1,713
620
NR
NR
NR
NR
NR
NR
18.2
NR
Texas
Site(a)
273
82
1,007
26.5
NR
579
1,009
1,181
NR
6.71
4,186
NR
NR
NR
NR
NR
NR
NR
28.6
NR
Kingsville Dome
Site, Texas(b)
560
92
800
31
NR
619
919
1,660
23.5
6.82
4,640
507
0.016
0.02
1.7
32
293
0.104
29.0
0.01
Typical
Chemistry(c)
100-350
10-50
500-1,600
25-250
0-500
800-1,500
250-1,800
100-1,200
25-50
7-9
1,500-5,500
NR
NR
NR
NR
NR
500
NR
50-250
NR
Typical Lixiviant
Chemistry Range(d)
< 20-500
< 3-100
< 400-6,000
< 15-300
< 0.5-2,500
< 400-5,000
< 200-5,000
< 400-5,000
NR
< 6.5-10.5
< 1,650-12,000
NR
NR
NR
NR
NR
NR
NR
< 0.008-424
< 0.006-56
All units are mg/L unless otherwise noted.
NR = Not reported
^ Deutschetal. 1985
b Schramke et al. 2009
(c) Pelizza2008
(d) NRC2009b
4-1
-------
Additional information on lixiviant concentrations is included in Table 4-2: (CNWRA 2001).
These are the maximum quoted concentrations based on a survey of licensing documents.
Table 4-2: Highest Observed Concentrations in Pregnant Lixiviants
based on a Survey of Licensing Documents
Element/Isotope
Ra-226
Lead
Uranium
Concentration
3,400 pCi/L
0.01 mg/L
250 mg/L
Based on this information, lixiviant concentrations were established as follows.
Lead Concentration
The maximum amount of lead reported in lixiviants is 0.01 mg/L (Table 4-2).
Uranium Concentration
The maximum amount of uranium reported in lixiviants is 424 mg/L (Table 4-1). This value can
be traced to NRC 1989, but that reference is silent on the source of the data, noting only that the
"values represent the concentration ranges that could be found in barren lixiviant and pregnant
lixiviant and would include the concentrations normally found in 'injection fluids'" (NRC 1989,
Table 3.4.01). Use of 424 mg/L as an upper bound on the uranium concentration thus lacks
credibility. The values in Column 5 of Table 4-1 are from Pelizza 2008, but Pelizza also does not
provide the source of the data. It likely came from NUREG-1508 (NRC 1997), which also lists
the range for uranium from 50 to 250 mg/L. According to Table 2.1 of NUREG-1508, these data
were obtained from HRI1993 and are based on test data and operational licensing experience.
This is presumably from the same source as cited in Table 4-2 above. CNWRA 2001 cites
60 mg/L as a typical concentration for pregnant lixiviant. Another source of information is spill
incidents in 1999 at the Smith Ranch involving injection fluids with a natural uranium
concentration of 2.7E-06 |iCi/ml or 3.9 mg/L, and extraction fluids with a concentration of 5.3E-
05 nCi/ml or 77 mg/L (NRC 2000). Additionally, NRC reported that since June 1997, uranium
releases in spills at the Smith Ranch-Highland Uranium Project ranged from 0.7 to 152 mg/L,
with about 70% of the releases below 10 mg/L (NRC 2007). Based on this information, we have
selected a uranium concentration of 150 mg/L as a reasonable upper-end value for calculating
doses and risks.
Radium Concentration
The highest observed Ra-226 concentration in lixiviants is 3,400 pC/L (Table 4-2). Based on the
specific activity of 1 Ci/g for Ra-226, this is equivalent to 3.4E-06 mg/L. This value is based on
the Ra-226 concentration in production fluids at the Smith Ranch Facility in Converse County,
Wyoming (NRC 2000). The Ra-226 concentration in injection fluids was similar (i.e.,
3,300 pCi/L); radium is not removed when uranium is extracted from the pregnant lixiviant and
is returned underground with the injection solutions.
4-2
-------
Thorium Concentration
As noted by the NRC in its Standard Review Plan for ISLs (NRC 2003, Section 5.7.8.3):
many licensees have decided not to sample for Th-230; Th-230 is a daughter
product from the decay ofuranium-238, and studies have shown that it is
mobilized by bicarbonate-laden leaching solutions. However, studies have also
shown that after restoration, thorium in the ground-water will not remain in
solution, because the chemistry of thorium causes it to precipitate and chemically
react with the rock matrix (Hem, 1985). As a result of its low solubility in natural
waters, thorium is found in only trace concentrations. Additionally, chemical tests
for thorium are expensive, and are not commonly included in water analyses at in
situ leach facilities. This example concerning Th-230 demonstrates an acceptable
technical basis for excluding Th-230 from the list of sampled constituents.
In the absence of detailed geochemical modeling to support the immobility of thorium in ground
water, we have selected a value of 640 pCi/L (3.05E-05 mg/L) for the Th-230 source term. This
value was measured at the Irigaray Solution Mining Project in the post-leaching ground water
prior to restoration (NRC 1978, Table 5.1).
4.2.2 Selection of Kd Values
Selection of KdS was based on selecting low-end values to conservatively establish the time to
reach peak dose at a down-gradient well. It should be noted that the Kd value affects the time at
which the peak dose occurs, but not its magnitude in the absence of significant radioactive decay.
Considerations in selecting the Kd values are described below.
LeadKd
EPA 1999b provides a thorough survey of the adsorption behavior of lead in soil studies and
identifies the factors that influence the behavior of lead. The pH of the aqueous phase had an
important effect on lead sorption, which increases over the pH range of 4-11. Table 4-3
summarizes lead Kd values as a function of pH and aqueous lead concentrations from EPA
1999b. It is probable that in the ISL environment, lead will derive from the uranium series decay
chain and will be present in low concentrations.
4-3
-------
Table 4-3: Estimated Range of Kd Values for Lead as a Function of Soil pH, and
Equilibrium Lead Concentrations
Equilibrium Lead Concentration
(micro g/1)
01 ft Q
1 ft Q Q
1 ft QQ Q
1 flfl Oflfl
K., (m\M
f^d V''"g^
Minimum
Maximum
Minimum
Maximum
Minimum
Maximum
Minimum
Maximum
4.0-6.3
940
8,650
420
4,000
190
1,850
150
860
Soil pH
6.4-8.7
4,360
23,270
1,950
10,760
900
4,970
710
2,300
8.8-11.0
11,520
44,580
5,160
20,620
2,380
9,530
1,880
4,410
Based on this information, we selected a minimum Kd for 10 |ig/L of lead in solution of 900 ml/g
for pH 6.4-8.7. This pH range is consistent with observed values at ISL facilities (Davis and
Curtis 2007, Table 3 and 4).
With a Kd of 900 ml/g for lead and conservative hydraulic data from existing ISL facilities (see
Table 4-8 and Figure 4-1), it would take the isotopes of lead initially present in the ground water
approximately 260,000 years to reach the nearest well located at a distance of 528 ft from the
facility. Since the half-life of Pb-210 is 22.3 years, any Pb-210 contamination initially present in
the lixiviant would completely decay away before it could arrive at the receptor well.
Furthermore, as described in Section 3.2.2, any in-growth of Pb-210 from the decay of Ra-226
has been accounted for in this analysis by adding the Pb-210 dose and risk factors (as well as the
dose and risk factors for other short-lived Ra-226 progeny) to the Ra-226 dose and risk factors.
Thus, the Ra-226 doses and risks presented in this report include the contribution from Pb-210
and the other Ra-226 short-lived progeny.
Uranium Kd
Aqueous uranium and its complexes sorb onto clays, organics, and iron oxides (EPA 1999b).
Uranium sorption by soils generally reaches a maximum in the pH range from pH 5 to 8 (EPA
1999b). Higher ionic-strength solutions or the presence of carbonate ions tend to decrease
uranium(VI) sorption. Uranium can also be attenuated in ground water through co-precipitation
reactions with metal oxyhydroxides, such as iron hydroxide.
Table 4-4 provides a uranium Kd look-up table from EPA 2004. The general trend in uranium Kd
values as a function of pH is that adsorption is low at pH values of 3 or less, increases rapidly
from pH of 3 to 5, reaches a maximum between pH of 5 and 8, and then decreases with
increasing pH greater than 8. The decrease in absorption at high pH is actually related to the
presence of dissolved carbonate. At near- and above-neutral pH conditions, dissolved U(VI)
forms strong anionic uranyl-carbonate complexes with dissolved carbonate, making it less likely
to adsorb to the surface-charged soil minerals (EPA 2004).
4-4
-------
Table 4-4: Estimated Range
Values for Uranium based on pH
Kd
(ml/g)
Minimum
Maximum
PH
3
<1
32
4
0.4
5,000
5
25
160,000
6
100
1,000,000
7
63
630,000
8
0.4
250,000
9
<1
7,900
10
<1
5
Source: EPA 1999b
As documented in EPA 1999b (Section J.40):
Under oxidizing conditions atpH values greater than 6, their derived Kd values
were approximately 100 ml/g. At high concentrations of dissolved carbonate, and
pH values greater than 6, the Kd values for uranium decrease considerably.
Based on these considerations, we have selected a value of 0.4 ml/g (the minimum value at pH 8
in Table 4-4).
Radium Kd
Radium, an alkaline earth element, is generally relatively immobile, but can be mobilized under
some conditions. Radium-226 and radium-228 are present in uranium roll-front deposits, because
of the decay of uranium-238 and thorium-232, respectively. Ground water radium concentrations
commonly are elevated in the ore zone relative to the background levels present immediately up-
gradient and down-gradient of the ore (Hoy 2006). In general, radium adsorption on mineral
surfaces increases with increasing pH. For iron oxides, the increase in adsorption begins around
pH of 6 to 8 and reaches a maximum of around 10 or less (EPA 2004). Radium can be attenuated
by adsorption onto clays. Radium is also strongly adsorbed to mineral oxides, especially at near-
neutral and alkaline pH conditions (EPA 2004).
Compared to most other radionuclides, very limited data are available on radium sorption,
particularly Kd values. Moreover, EPA (2004) states that any data indicating high radium
adsorption on geologic materials should be viewed carefully, as (Ba, Ra) S04 co-precipitation
may have occurred during the measurements. However, from an attenuation standpoint, the
amount of radium that is removed from ground water by adsorption versus precipitation is
largely irrelevant. EPA (2004) presents ranges of Kd values by soil type, as shown in Table 4-5.
As noted by EPA (2004), "radium is readily adsorbed to clays and mineral oxides present in
soils, especially at near neutral and alkaline pH conditions."
Table 4-5: Radium Ka Values by Soil Type
Soil Type
Sand
Silt
Clay
Organic
Kj Values (ml/g)
Geometric Mean
500
36,000
9,100
2,400
Number of Observations
3
3
8
1
Range
57-21,000
1,262 - 530,000
696 - 56,000
None Listed
Source: EPA 2004
4-5
-------
As can be seen from Table 4-5, a minimum Kd value of 57 ml/g is cited for radium in sand. NRC
recommends a value of 100 ml/g for alkaline (cementitious) environments (NRC 1998). This
seems more relevant to alkaline ISL environments and has been selected for calculating
retardation factors here.
Thorium Kd
NRC (1998) recommended a Kd value of 500 ml/g for thorium in a Type III environment, which
is equivalent to highly weathered cement. EPA (1999b, Table 5-15) quotes a minimum value of
20 ml/g for Th concentrations of <10"9 M (2.3E-07 g/L) and pH of 8 to 10. For the pH range of 5
to 8, EPA quotes a minimum value of 1,700 ml/g. For the analyses presented here, the
intermediate value of 500 ml/g was selected.
4.2.3 Recommendations for Kd and Radionuclide Source Term
Recommended Kd values and source concentrations for radionuclides are summarized in Table
4-6. It is recognized that, if leakage is from an injection well as contrasted to an extraction well,
the concentrations, particularly uranium, will be lower than cited in Table 4-6. If lixiviant
concentrations at actual sites are different than those assumed in Table 4-6, the results can be
scaled linearly from those used here.
Table 4-6: Summary of Ka Values and Lixiviant
Concentrations Used in Dose and Risk Calculations
Radionuclide
Unat
Th-230
Ra-226
Pb-210
Kd (ml/g)
0.4
500
100
900
Lixiviant Concentration
(mg/L)
150
3.05E-05
3.4E-06
0.01
Radionuclide-specific retardation factors may be calculated using the following equation:
R,
R!
Pb
Pt
= 1+-
4-1
Pt
= Retardation factor of radionuclide /' in the saturated zone
= Saturated zone soil bulk density
= Radionuclide / distribution coefficient in the saturated zone
= Total porosity of the saturated zone (dimensionless)
(g/cm3)
(cm3/g)
Radionuclide-specific retardation factors were calculated using equation 4-1, the above KdS, a
soil bulk density (pb) of 2 g/cc, and a total porosity (pt) of 0.3.
4-6
-------
4.3 Dose and Risk Calculations
4.3.1 Limiting Doses and Risks
For illustrative purposes, limiting concentrations at the receptor well are derived for an assumed
dose limit of 15 mrem/yr and a lifetime risk limit of 10~4 LCF12, or assuming a 70-year life
expectancy, an annual risk limit of 1.4 x 10~6 LCF/yr. The limiting radionuclide concentrations
shown in Table 4-7 were derived using the adult, mean pathway dose and risk conversion factors
(PDCFs and PRCFs) derived in Chapter 3. Based on the approach taken for the CAP88 computer
program, the adult PDCFs and PRCFs were used to calculate the Table 4-7 limiting
concentrations. The limiting concentrations for Ra-226 in Table 4-7 are shown both with and
without progeny. Limits for one daughter product, Pb-210 with a 22.3-year half-life, are also
shown separately.
Table 4-7: Radionuclide Well Limiting Concentrations
Nuclide
U-238+P
U-234
U-natural
Th-230
Ra-226
Ra-226 + P
Pb-210+P
15 mrem/yr Dose Limit
(pCi/L)
97
93
95
25
1.2
2.4
97
(mg/L)
2.8E-01
1.5E-05
1.4E-01
1.2E-06
1.2E-09
3.1E-11
2.8E-01
10 4 LCF Lifetime Risk Limit
(pCi/L)
63
65
64
47
0.6
1.3
63
(mg/L)
1.8E-01
l.OE-05
9.4E-02
2.2E-06
6.3E-10
1.7E-11
1.8E-01
U-238+P included progeny:Th-234
Ra-226+P included progeny:Pb-214, Bi-214, Pb-210, Bi-210, Po-210
Pb-210+P included progeny: Bi-210, Po-210
As Table 4-7 demonstrates, a lifetime risk limit of 10~4 LCF is slightly more restrictive than a
dose limit of 15 mrem/yr for uranium, Ra-226, and Pb-210, while the dose limit is slightly more
restrictive for Th-230. However, for all of the radionuclides considered, both dose and risk
limiting concentrations are within a factor of two.
4.3.2 Doses and Risks from Excursion Scenarios
As described in Section 2.6.2, 63 excursion scenarios were analyzed in which some lixiviant
leaked down-gradient from various injection/extraction pumping patterns. The ground water
modeling studies described in Chapter 2 included broad ranges of selected hydraulic parameters
to assist in evaluating the sensitivity of receptor well concentrations to parameter changes. While
the selected ranges for individual parameters were based on literature reviews, some
combinations of hydraulic parameters may not exist in actual licensed ISL facilities. For
example, ground water velocity is defined as the product of hydraulic conductivity and hydraulic
12 A lifetime cancer morbidity risk limit of 10"4 was specified in the National Contingency Plan (55 Federal
Register 8665-8865, March 8, 1990). In this example the cancer morbidity risk limit is being applied to the risk of
latent cancer fatalities (LCF). For the radionuclides analyzed in this study, the relationship between cancer morbidity
and mortality is discussed in Section 3.2.2.
4-7
-------
gradient divided by the effective porosity. In some of the modeling scenarios, both high
conductivity and high gradient were included, which could result in unrealistically high
velocities. To focus on doses and risks from those scenarios that are most representative of actual
ISL conditions, we obtained data from several proposed and operating ISL facilities, as
summarized in Table 4-8. From these data, we constructed the cumulative distribution for the
product of hydraulic gradient times hydraulic conductivity function shown in Figure 4-1. It can
be seen that all of the values fall below a limit of 0.13 (ft/day). Based on this limit, we developed
doses and risks for those modeled excursion scenarios not exceeding this limit.
Table 4-8: ISL Site Hydraulic Data
Site / Identifier
Crow Butte, NE
Goliad County, TX
Irigaray, WY
Kingsville Dome, TX
Lost Creek, WY
Major
Minor
PBL-1
PBL-2
PBL-3
PBL-5
Multi-well
Single well
Moore Ranch, Campbell County WY
Nichols Ranch, WY
North Butte, WY
A sand
F sand
A sand
B and C sand
Hydraulic
Conductivity
(ft/day)
9.11
3.5
1.55
0.896
24.48
14.112
20.16
12.672
1.8 to 4.4
4.4 to 11.7
5.36
0.5
0.6
8.79
8.43
Hydraulic
Gradient
0.0004
0.0009
0.033
0.033
0.005
0.0009
0.005
0.0009
0.003
0.006
0.004
0.0033
0.005
0.015
0.0061
Reference
Crow Butte 1995
(Table 2.7-6 and p.2-13)
DBS&A 2007
(Table 1 and Sec. 2.5)
NRC 1978 (p. 2-16)
Rice 2006
(Tables A-2, A-3, A-4)
Lost Creek 2007
(Tables 2.7-7 and 2.7-9)
NRC 2009a
(Sec. 3.5.2.3)
NRC 2009c
(Sec. 3.5.2.3)
PRI 2006
(Table 10. land p. 10-13)
4-8
-------
0.02 0.04 0.06 0 08 0.10 0.12 0.14
Hydraulic Conductivity (ft/day) times Hydraulic Gradient
Figure 4-1: Cumulative Distribution of ISL Site Hydraulic Data
Thirty-seven (37) analyzed excursion scenarios met Figure 4-1 hydraulic data criteria and were
included in the dose and risk analysis. Figure 4-2 shows the dose to an adult from uranium for
scenario 6E at each of the four assumed down-gradient receptor well locations. From Figure 4-2,
it is apparent that as the down-gradient distance to the well increases, two things occur; (1) the
dose to the receptor decreases, and (2) the time after the release when the peak dose occurs
increases. As an example, from Figure 4-2, the dose to an adult from a well at 528 ft is about
7,600 mrem/yr and occurs at about 20 years, whereas the dose to an adult from a well at 3,480 ft
is about 610 mrem/yr and occurs at about 124 years. Unless otherwise specified, all of the
remaining doses and risks discussed in this section are based on the assumption that the receptor
well is located at a distance of 528 ft down-gradient.
Figure 4-2 also illustrates an important factor regarding wellfield restoration. If lixiviant has
escaped from the wellfield during operations, it will not arrive at a well 528-ft away for at least
20 years, depending on the actual Kd. Thus, the peak uranium dose at the down-gradient well is
likely to occur long after restoration of the wellfield has been completed, based on current
practices.
4-9
-------
80 100
Time (years)
' GO
I BU
Figure 4-2: Excursion Scenario 6E Peak Dose Arrival Time, Kd = 0.4 ml/g
Figure 4-3 shows the adult mean doses at the time of maximum concentration that were
calculated for each scenario and radionuclide. (Note that some of the 37 scenarios and
radionuclide maximum doses were lower than the 0.01 mrem/yr cutoff and are not shown in
Figure 4-3.) All of the doses shown in Figure 4-3 are for a receptor located at the nearest well, at
a distance of 528 ft. In Figure 4-3 and all subsequent tables and figures in Chapter 4, radiological
decay is not explicitly included; however, whether specifically noted or not, contribution of the
progeny to the dose/risk is included as described in Section 3.2.2.
1 E+05 -
B
g
a
<
|
* U-nat
A * Ra-226
>: * :
t
;
.
+ 15 mrem/yr
•
*
:
• *
f m
•
1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06
Time (years)
Figure 4-3: Excursion Scenario Dose Results versus Time
4-10
-------
In Figure 4-3, the receptor doses due to uranium occur first, due to the assumed low value of the
uranium Kd. Also, the uranium doses are received in two distinct groupings—the first around
30 years and the second around 120 years. This grouping tendency is due to the fact that the
product of the hydraulic gradient and hydraulic conductivity for 36 of the 37 scenarios analyzed
was either 0.01 or 0.1 ft/day (see Figure 4-4). This same pattern is shown in Figure 4-3 for the
Th-230 and Ra-226 (+ progeny) doses. Because each radionuclide arrives at the receptor well at
a different time, due to their different Kd values, it is not necessary to sum the exposure from
different radionuclides. Table 4-9 shows the maximum calculated dose for each radionuclide
from all 37 excursion scenarios that were analyzed. Clearly, uranium and Ra-226 (+ progeny) are
the significant contributors to dose and risk.
Table 4-9: Excursion Scenario Maximum Doses and Risks -
Mean Adult
Nuclide
Unat
Th-230
Ra-226+ P
Dose (mrem/yr)
l.OE+04
2.4E+02
2.8E+04
Risk (LCF/yr)
1.4E-03
1.2E-05
4.8E-03
1 E+04 -
^
O)
o>
tf>
O
O
<
1
1 E-01 -
« U-nat
•Th-230 f
Ra-226
A
*
*
!
ft
1 5 mrem/yr ^
~* •
! ;
8
A •
*
1.E-04 1.E-03 1.E-02 1 E-01 1.E+00
Hydraulic Gradientx Hydraulic Conductivity (ft/day)
Figure 4-4: Excursion Scenario Dose Results versus Hydraulic Data
Figure 4-4 shows the same dose results as Figure 4-3, except as a function of their hydraulic
data, instead of as a function of time. As indicated above, Figure 4-4 shows that 36 of the 37
excursion scenarios analyzed have the product of their hydraulic gradient and hydraulic
conductivity equal to either 0.01 or 0.1 ft/day; in one scenario, the product was 0.001 ft/day.
Figure 4-1 shows that a hydraulic data product of 0.1 ft/day or more occurs in only 10% of the
actual ISL sites, whereas a hydraulic data product of 0.01 ft/day or less occurs in about 43% of
the actual ISL sites. Thus, although very high doses were calculated for some of the excursion
4-11
-------
scenarios, many of the scenarios with high doses are at hypothetical sites with hydraulic
properties that occur very infrequently at actual ISL sites.
Figure 4-5 is similar to Figure 4-3, except that it shows the calculated risks. The above
discussion of the dose results also applies to the risk results, and is not repeated.
1
o
Mean Adult Risk (
l m i
3 b i
•j m t
» U-nat
J. * BTh-230
i Ra-226
i
A
2
J
I.
A A
»
• 1 0-4 LC F/l ifetim e = 1 A x 1 0-8 LC Ry r •
_ _ _* 1 f .
.
• .
*
A •
• A B
:
•
1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E
Time (years)
toe
Figure 4-5: Excursion Scenario Risk Results versus Time
When viewing Figure 4-5, it should be remembered that a lifetime risk limit of 10~4 LCF
corresponds to an annual risk of about 1.4x 10~6 LCF/yr.
Figure 4-2 shows the annual doses that result at the various assumed well locations due to the
passing of the uranium contamination plume. In developing Figure 4-2, it was assumed that the
Kd for uranium was 0.4 ml/g, as described in Section 4.2.2. However, Section 4.2.2 also pointed
out that there is a great deal of variability in the possible uranium Kd, with Table 4-4 showing
that potentially the uranium Kd could be in the range of 100 to 1,000,000 ml/g.
Figure 4-6 shows the effect on the dose by arbitrarily increasing the uranium Kd by a order of
magnitude from 0.4 to 4 ml/g. When comparing the Figure 4-6 doses to the Figure 4-2 doses,
notice that the only thing that changes is the time when the plume arrives at each well location,
but that the magnitude of the doses received remains unaffected by the choice of the uranium Kd.
This effect is due to the very long half-life of uranium-238 (i.e., 4.47><109 years), which results in
virtually no additional radiological decay during its longer travel time to the receptor wells.
Because Th-230 and Ra-226 also have long half-lives (i.e., 75,380 and 1,600 years, respectively),
the doses due to Th-230 and Ra-226 would likewise be unaffected by changes to their Kd values.
4-12
-------
^r
w
O i F+01 -
Q ' C Ul
±±
2
•o
S 1 E+00 -
I I
Well: 528 ft
Well: 856 ft
— Well: 1840 ft
Well: 3480 ft
1 5 mrem/yr
T
1
0 100 200 300 400 500
Time (years)
600 700
(\
f
\
800 900
Figure 4-6: Excursion Scenario 6E Peak Dose Arrival Time, Ka = 4.0 ml/g
4.3.3 Doses and Risks from Surface Leak Scenarios
As described in Chapter 2, 24 scenarios were analyzed that postulated the spill of lixiviant onto
the ground, with subsequent leakage into the ground water and transport to an offsite receptor
well. As explained in Chapter 2, three types of surface leak scenarios were analyzed:
(1) catastrophic spills ranging from 100,000 to 200,000 gallons, (2) a slow leak of 1 to 2 gpm for
a period of 3 years, and (3) leaks varying from 1 to 45 gpm over a 28-day period. Figure 4-7 and
Figure 4-8 show the calculated doses and risks for all of the 24 leak scenarios, respectively, and
Table 4-10 shows the maximum and minimum dose and risks from Figure 4-7 and Figure 4-8.
All of the doses and risks shown are for a receptor located at the nearest well at a distance of
328 ft. Notice on Figure 4-7 and Figure 4-8 that the first uranium doses are received earlier for
some of the leak scenarios than for any of the excursion scenarios. This is due to the fact that for
three leak scenarios, the product of their hydraulic gradient and hydraulic conductivity was equal
to 1.0 ft/day. This higher hydraulic data product was felt to be reasonable for the leak scenarios,
because it is believed that the leak could travel in regions near the ground surface that have both
a high gradient and a high conductivity (UT 2010). Thus, for the surface leak scenarios, the
uranium contamination plume could arrive at a receptor well very soon after the spill occurs
(e.g., within 2 to 3 years).
4-13
-------
¥
0
a
±i
B
to
i
i Ll-nat
• Th-230
* A Ra-226
»I
! «
, «» *
t *
* *
:
15 mrem/yr
•
! i
1 • •
•
1.E+00 LE-i-01 1.E+02 1.E-HD3 1.E+04 1.E+05 1.E+06
Time (years)
Figure 4-7: Surface Leak Scenario Risk Results versus Time
I
X
CA
DC
i
re
i
U-nat
• Th-230
»
« Ra-226
! t
' • i
*
1 CH LC F/lifetime = 1 4 « 1 0-B LC F/vr •
"! •
•
! i
•
•
1.E+00 1.E+O1 1.E+02 1.E+03 1.E+O4 1.E+05 1.E+06
Time (years)
Figure 4-8: Surface Leak Scenario Risk Results versus Time
Table 4-10: Surface Leak Scenario Doses and Risks - Mean Adult
Unat
Th-230
Ra-226 + P
Dose (mrem/yr)
Minimum
3.2E+01
7.6E-01
8.7E+01
Maximum
1.7E+03
4.0E+01
4.6E+03
Risk (LCF/yr)
Minimum
4.5E-06
3.9E-08
1.5E-05
Maximum
2.4E-04
2.0E-06
7.9E-04
4-14
-------
Leak scenarios 19 through 24 assumed that leaks of different magnitudes occurred for 28 days.
Figure 4-9 shows the doses due to uranium as a function of the various leak rates at each of the
four assumed receptor well locations. As Figure 4-9 shows, there is almost a linear relationship
between the amount of leaking lixiviant and the doses received at the receptor wells.
1,000
15 20 25 30
Leak Rate (gpm)
Figure 4-9: 28 Day Surface Leak Scenario Doses from Uranium
4.4 Risks to Non-standard Receptors
All of the doses and risks discussed in Section 4.3 were based on the receptor being an adult with
mean usage rates, i.e., the adult mean PDCFs and PRCFs from Tables 3-10 and 3-11,
respectively. This was the approach taken in CAP88 to demonstrate compliance with 40 CFR
Part 61, Subpart H. However, it is recognized that there may be individuals living near the ISL
facility who are not adults, or who have usage rates that significantly deviate from the mean. To
address those individuals, this section describes the results of calculations that have been
performed for non-standard dose and risk receptors.
In Chapter 3, PDCFs and PRCFs were developed for a number of non-standard receptors,
including:
• 90t Percentile Adult - an individual who has increased usage rates, such that his/her
PDCFs and PRCFs are at the 90th percentile, as calculated from the usage rate
distributions described in Chapter 3. Only 10% of adults would be expected to have
larger PDCFs and PRCFs than this individual.
• Mean Teenager - an individual who is between the age of about 13 and 19 years old, with
mean usage rates.
• Mean Child - an individual who is between the age of about 1 and 12 years old, with
mean usage rates.
• Mean Infant - an individual who is less than 1-year old, with mean usage rates.
4-15
-------
• Native American - an adult individual with exposure pathways and usage rates that are
typical for a Native American (e.g., exposure during a Sweat Lodge Ceremony).
Refer to Section 3.4 for more information and discussion of the Native American exposure
pathways and usage rates, and to Section 3.2 for more information regarding the other non-
standard receptors. The Native American PDCFs and PRCFs are given in Table 3-20, while
Tables 3-10 and 3-11 give the PDCFs and PRCFs for the other non-standard receptors,
respectively.
Table 4-11 and Table 4-12 show the calculated doses and risks to the non-standard receptors for
the excursion and leak scenarios, respectively. Notice that for the excursion scenarios, only the
maximum dose and risks are presented in Table 4-11, while in Table 4-12, both the maximum
and minimum doses and risks are presented. This was done because the minimum doses and
risks for some of the excursion scenarios are negligibly small. For Table 4-11 and Table 4-12, it
is apparent that the Mean Infant is the recipient of the largest calculated doses and risks.
Table 4-11: Excursion Scenario Non-standard Receptor Doses and Risks"
Receptor
90th Percentile Adult
Mean Teenager
Mean Child
Mean Infant
Native American
Nuclide
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Maximum
Dose (mrem/yr)
1.7E+04
4.4E+02
4.7E+04
9.6E+03
1.5E+02
4.2E+04
8.7E+03
1.2E+02
5.6E+04
2.7E+04
1.7E+03
1.5E+05
2.5E+04
7.8E+02
5.7E+04
Risk (LCF/yr)
2.5E-03
2.2E-05
8.2E-03
2.9E-03
1.8E-05
8.2E-03
5.1E-03
1.9E-05
1.5E-02
8.5E-03
5.3E-05
1.6E-02
4.3E-03
4.0E-05
l.OE-02
a - Doses for Ra-226 include progeny
Table 4-12: Leak Scenario Non-standard Receptor Doses and Risks"
Receptor
90th Percentile Adult
Mean Teenager
Mean Child
Mean Infant
Nuclide
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Unat
Th-230
Ra-226
Unat
Dose (mrem/yr)
Minimum
5.4E+01
1.4E+00
1.5E+02
3.0E+01
4.5E-01
1.3E+02
2.7E+01
3.6E-01
1.7E+02
8.4E+01
Maximum
2.9E+03
7.2E+01
7.8E+03
1.6E+03
2.4E+01
7.0E+03
1.4E+03
1.9E+01
9.2E+03
4.4E+03
Risk (LCF/yr)
Minimum
7.7E-06
7.0E-08
2.6E-05
9.1E-06
5.7E-08
2.6E-05
1.6E-05
5.9E-08
4.8E-05
2.7E-05
Maximum
4.1E-04
3.7E-06
1.4E-03
4.8E-04
3.0E-06
1.3E-03
8.4E-04
3.1E-06
2.6E-03
1.4E-03
4-16
-------
Table 4-12: Leak Scenario Non-standard Receptor Doses and Risks"
Receptor
Native American
Nuclide
Th-230
Ra-226
Unat
Th-230
Ra-226
Dose (mrem/yr)
Minimum
5.4E+00
4.7E+02
7.8E+01
2.4E+00
1.8E+02
Maximum
2.8E+02
2.5E+04
4.1E+03
1.3E+02
9.4E+03
Risk (LCF/yr)
Minimum
1.6E-07
5.1E-05
1.3E-05
1.2E-07
3.2E-05
Maximum
8.7E-06
2.7E-03
7.1E-04
6.5E-06
1.7E-03
a - Doses for Ra-226 include progeny
The doses and risks to the non-standard receptors in Table 4-11 and Table 4-12 can be compared
to the dose and risks presented in Table 4-9 for the excursion scenarios and Table 4-10 for the
leak scenarios. This comparison shows that the Mean Infant doses and risks are about a factor of
three to eight times larger than the standard receptor doses and risks. The dose and risk ratios of
the other non-standard receptors to the standard receptor are less than for the Mean Infant. For
example, the Native American has a calculated maximum uranium risk that is three times greater
than the standard receptor's uranium risk for both excursion and leak scenarios. Also, the Native
American's maximum Ra-226 dose is two times greater than the standard receptor's Ra-226
dose.
Finally, in Chapter 3, PDCFs and PRCFs were developed for a number of potential receptors
who have not been included in this analysis, e.g., 90* Percentile Infant. If it is desired to include
any of these individuals, that can be accomplished by simply multiplying the results shown in
Table 4-11 and/or Table 4-12 by the ratio of the PDCFs or PRCFs, whichever is appropriate.
4.5 Non-Radiological Risks
In addition to radiological risks, the lixiviant used to leach the uranium from the ore will contain
other potentially hazardous constituents. A screening calculation has been performed in order to
obtain an indication as to how hazardous these other constituents in the lixiviant may be to an
individual located at the nearest receptor well. The screening calculation consisted of comparing
the well water concentration of the potentially hazardous constituents to the Safe Drinking Water
Act concentration limits for that constituent.
Maximum lixiviant concentrations were obtained from NUREG/CR-6733 (CNWRA 2001,
Table 4-7), with additional lixiviant concentrations from NUREG/CR-6970 for Highlands and
Crow Butte (Davis and Curtis, 2007, Tables 3 and 5), as well as Smith Ranch Mine Unit B
(MUB) (Power Resources 2004, Table 4), and Christensen MU2 and MU3 (COGEMA, 2008,
MU2 Table 5.1 and MU3 Table 5.1). Table 4-13 shows all of the lixiviant concentrations at the
end of mining that were collected.
Table 4-13 also provides the maximum concentration limit (MCL) for each of the potentially
hazardous constituents. An MCL is the legal threshold limit on the amount of a substance that is
allowed in public water systems under the Safe Drinking Water Act. When an MCL is
unavailable, the Lifetime Health Advisory (HA) concentration is provided. The Lifetime HA is
the concentration of a chemical in drinking water that is not expected to cause any adverse
noncarcinogenic effects for a lifetime of exposure. Both the MCLs and the Lifetime HAs were
4-17
-------
obtained from the 2011 Edition of the Drinking Water Standards and Health Advisories (EPA
2011 a).
Table 4-13: Highest Contaminant Levels in Pregnant Lixiviant (mg/L)
Contaminant
Ammonium
Arsenic
Barium
Boron
Cadmium
Chloride
Chromium
Copper
Fluoride
Lead
Manganese
Mercury
Molybdenum
Nickel
Nitrate
Nitrite
Selenium
Uranium
Zinc
NUREG/C
R-6733
N.P.
0.3
0.6
0.2
0.01
1800
0.03
0.04
1
0.01
6
O.0001
62
0.09
1
N.P.
5
250
N.P.
Crow
Butte
0.37
0.002
0.1
0.93
0.006
204
<0.03
0.017
0.69
0.031
0.11
0.001
0.069
0.034
0.05
0.01
0.003
0.092
0.036
Highland
0.1
0.001
0.1
0.1
0.01
4.7
0.05
0.01
0.2
0.05
0.03
0.001
0.1
0.05
0
0
0.001
0.05
0.01
Smith
Ranch
0.52
0.008
0.1
0.1
0.01
232
0.1
0.01
0.1
0.1
0.9
0
0.1
0.07
0.3
0.1
0.806
22.3
0.11
Christensen
MU2
0.52
0.12
0.1
0.1
0.01
122.9
0.05
0.01
0.1
0.05
0.66
0.001
0.1
0.12
0.22
0.12
6.33
11.75
0.05
Christensen
MU3
1.14
0.02
0.1
0.1
0.01
155.4
0.05
0.01
0.1
0.05
0.69
0.001
0.1
0.05
0.1
0.1
4.34
15.58
0.01
Limit
30
0.01
2
6
0.005
4
0.1
1.3
4
0.015
0.3
0.002
0.04
0.1
1
10
0.05
0.03
2
Type
Life -time
MCL
MCL
Life -time
MCL
MCL
MCL
MCL
MCL
MCL
Life -time
MCL
Life -time
Life -time
MCL
MCL
MCL
MCL
Life -time
For each potentially hazardous constituent, Table 4-14 first shows the maximum lixiviant
concentration (mg/L) from Table 4-13, and then shows the MCL (or Lifetime HA) divided by the
maximum concentration. If the quotient is 1 or greater than that constituent is not at a hazardous
concentration in the ore zone or at the nearest receptor well.
Finally, as stated previously (Section 4.3.2) there were 37 excursion scenarios that met the
hydraulic data criteria and were included in the dose and risk analysis. Each potentially
hazardous constituent concentration was compared to each analyzed scenario's dilution
coefficient at the receptor well. For each constituent Table 4-14 shows the number of scenarios
that have dilution coefficients which would reduce the concentration at the receptor well to
below the MCL (or Lifetime HA). For example, there are 11 analyzed scenarios resulting in dilution
coefficients that reduce the maximum molybdenum concentration at the receptor well to below the
molybdenum Lifetime HA. The final column of Table 4-14 simply shows the percentage of
scenarios that have constituent concentrations above the MCL (or Lifetime HA).
Table 4-14: Scenario Maximum Contaminant Levels Versus Limit
Contaminant
Nitrite
Copper
Ammonium
Maximum
(mg/L)
0.12
0.04
1.14
Source
Christensen MU2
NUREG/CR-6733
Christensen MU3
Limit
Maximum
83.3
32.5
26.3
Scenarios
Below Limit
at Receptor
Well
37
37
37
Scenarios
Above Limit
at Receptor
Well
0.0%
0.0%
0.0%
4-18
-------
Table 4-14: Scenario Maximum Contaminant Levels Versus Limit
Contaminant
Zinc
Boron
Fluoride
Barium
Mercury
Chromium
Nitrate
Nickel
Cadmium
Lead
Manganese
Arsenic
Selenium
Chloride
Molybdenum
Uranium
Maximum
(mg/L)
0.11
0.93
1
0.6
0.001
0.1
1
0.12
0.01
0.1
6
0.3
6.33
1800
62
250
Source
Smith Ranch
Crow Butte
NUREG/CR-6733
NUREG/CR-6733
Multiple Sources
Smith Ranch
NUREG/CR-6733
Christensen MU2
Multiple Sources
Smith Ranch
NUREG/CR-6733
NUREG/CR-6733
Christensen MU2
NUREG/CR-6733
NUREG/CR-6733
NUREG/CR-6733
Limit
Maximum
18.2
6.45
4.00
3.33
2.00
1.00
1.00
0.83
0.50
0.15
0.050
0.033
0.0079
0.0022
0.00065
0.00012
Scenarios
Below Limit
at Receptor
Well
37
37
37
37
37
37
37
37
36
33
24
24
17
17
11
7
Scenarios
Above Limit
at Receptor
Well
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
2.7%
10.8%
35.1%
35.1%
54.1%
54.1%
70.3%
81.1%
From Table 4-14 it can be seen that there are about five (5) lixiviant constituents (in addition to
uranium), which could have concentrations significantly above the MCL (or Lifetime HA). Also,
there are three (3) lixiviant constituents (i.e., nickel, cadmium, and lead), which have receptor
well concentration which are borderline hazardous. Finally, there are 10 lixiviant constituents
which are unlikely to be at hazardous concentrations, even within the ore zone.
4-19
-------
5.0 SUMMARY AND CONCLUSIONS
5.1 Ground Water Modeling Studies
Ground water modeling studies, described in Chapter 3, considered scenarios involving
excursions of contaminants from an ore-bearing aquifer and surface spills or leakage from
process components. Simulations of excursions within the ore zone aquifer evaluated the
sensitivity of the relative concentrations at down-gradient receptor wells to selected parameters.
Variables examined in these excursion simulations included:
• Well spacing (50, 150, and 250 ft)
• Hydraulic conductivity (1, 10, and 100 ft/day)
• Hydraulic gradient (0.001, 0.01, and 0.1 ft/ft)
• Pumping pattern (5-spot, multiple 5-spot, and 7-spot)
• Injection rates (7, 50, 150, and 500 gpm)
• Ore zone thickness (20 and 70 ft)
It is difficult to develop broad, general conclusions from these excursion simulations, because
results of comparisons designed into the modeling runs are at times counter-intuitive. In spite of
this difficulty, some conclusions are provided below. However, the reader is cautioned to read
the full text in Chapter 3 to understand the limitations of these conclusions.
• As expected, steeper hydraulic gradients result in shorter travel times. Furthermore, since
the pumping/injection wells are altering the regional hydraulic gradients, the arrival times
are not linearly scaled.
• An increase in hydraulic conductivity causes the contaminant plume to become more
dispersed and leads to lower relative peak concentrations.
• At higher regional gradients, wider well spacings provide better capture of the lixiviant.
At lower regional hydraulic gradients, however, better capture can be maintained at
smaller well spacings.
• The 7-spot well configuration results in lower relative peak concentrations for all of the
runs, as compared to a 5-spot pattern.
• The effect of pumping/injection on hydraulic gradients is strongly affected by the
transmissivity (i.e., hydraulic conductivity multiplied by thickness) of the geologic units.
The lower transmissivity results in shorter times to peak arrivals at the low and high
gradients, and a longer time at the medium gradient. These relationships are all related to
how the regional and localized gradients created by the pumping/injection interact to
form a capture zone. This also illustrates the complexity and need to understand the site-
specific geology and flow system, since the effects of the interactions are not always
intuitive.
• Comparison of a single 5-spot pattern with a multiple (25) 5-spot pattern shows that
relative concentrations at down-gradient receptor wells are lower with the multiple 5-spot
pattern. These results are explained by the larger capture zone that is created by the array
of pumping/extraction wells.
5-1
-------
As described in Chapter 3, 24 scenarios were analyzed, which postulated the spill of lixiviant
onto the ground, with subsequent leakage into the ground water and transport to an offsite
receptor well. The scenarios included (1) catastrophic spills ranging from 100,000 to
200,000 gallons, (2) a slow leak of 1 to 2 gpm for period of 3 years, and (3) leaks varying from
1 to 45 gpm over a 28-day period.
Transport from the mined aquifer to an overlying aquifer through an abandoned borehole that
was not properly cemented was also evaluated. This excursion scenario can result in significant
down-gradient leakage. This emphasizes the need to carefully cement and inspect abandoned
boreholes to insure their integrity.
5.2 Pathway Dose and Risk Conversion Factors
In Chapter 2, probabilistic dose and risk pathway conversion factors were developed for the
ingestion pathway for four age groups (infants, children, teens and adults) and five radionuclides
of importance in the ISL process (Ra-226+P, Pb-210, Th-230, U-234, and U-238). The
individual radionuclide-specific dose and risk conversion factors were based on FGR 13 and
supporting documentation. Contributions of the various components of the ingestion pathway to
dose and risk for adults are summarized in Table 5-1. Except for Ra-226, drinking water
accounts for about three-quarters of the total ingestion dose/risk. With Ra-226, ingestion of
vegetables is more significant than for the other radionuclides, accounting for about 35% of the
total ingestion dose/risk.
Table 5-1: Typical Pathway Contributions to the Adult PDCF and PRCF
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of Vegetables
Ingestion of Meat
Ingestion of Milk
Pb-210+P
78.3%
0.1%
15.7%
2.3%
3.5%
Ra-226+P
51.2%
0.2%
35.3%
2.9%
10.5%
Th-230
82.8%
0.3%
16.4%
0.3%
0.2%
U-234
76.2%
0.3%
16.1%
2.4%
5.0%
U-238+P
76.2%
0.3%
16.1%
2.4%
5.0%
Table 5-2 summarizes the pathway contributions from uranium exposure for each age group. The
shift from the dominance of the milk pathway to the drinking water pathway with increasing age
from child to adult is apparent.
Table 5-2: Typical Pathway Contributions to the PDCF and PRCF for
U-234 or U-238
Pathway - Adult
Ingestion of Drinking Water
Inadvertent Ingestion of Soil
Ingestion of Vegetables
Ingestion of Meat
Ingestion of Milk
Adult
76.2%
0.3%
16.1%
2.4%
5.0%
Teen
68.1%
0.5%
17.2%
3.4%
10.9%
Child
54.2%
1.4%
17.3%
2.8%
24.3%
Infant
81.2%
1.6%
10.2%
0.0%
6.9%
The pathway analysis for Native Americans was expanded from that of a standard receptor to
address exposures associated with sweat lodge rituals. This included additional drinking water
consumption, submersion in a cloud of contaminated vapor, and inhalation of contaminated
5-2
-------
vapor. The deterministic dose conversion factor for a Native American for natural uranium (i.e.,
U-234+U-238) is 4.05E-04 mrem/yr per pCi/m3, as compared to the 90th percentile dose
conversion factor for a standard adult receptor of 2.73E-04 mrem/yr per pCi/m3.
5.3 Dose/Risk Assessment for Modeled Scenarios
Doses and risks are calculated in Chapter 4 for some of the ground water flow and transport
scenarios modeled in Chapter 3. All modeling results in Chapter 3 excluded effects of retardation
and assumed a source term of 1 mg/L. In Chapter 4, source terms were selected based on
operating experience at ISL facilities, and included effects of retardation at down-gradient wells.
Distribution coefficients used to calculate the retardation factors were selected as minimum
reasonable values to reduce the travel time from the source to the receptor. As described in
Chapter 3, some combinations of hydrogeologic parameters, such as hydraulic conductivity and
hydraulic gradient, may lie outside the range expected at operating ISL sites, but were included
to examine the sensitivity of the modeling results to a range of parameters. In Chapter 4, doses
and risks were calculated only for those excursion scenarios with combinations of hydraulic
conductivity and hydraulic gradient that had been reported at operating ISL sites (i.e., gradient x
conductivity <0.13 ft/day). The product of hydraulic conductivity times hydraulic gradient was
used as a surrogate for ground water velocity (i.e., hydraulic conductivity x gradient + effective
porosity). An empirical cumulative distribution function was developed from the site data and
served to limit the range of parameters for which doses and risks for excursion scenarios were
calculated.
Prior to calculating doses and risks for the selected modeling scenarios, scoping calculations
were performed to illustrate the allowable radionuclide source term concentrations for
representative dose and risk limits. Limiting concentrations were derived for an assumed dose
limit of 15 mrem/yr and a lifetime risk limit of 10~4 latent cancer fatalities (LCF), or, assuming a
70-year life expectancy, an annual risk limit of 1.4 x 10~6 LCF/yr. The limiting radionuclide
concentrations shown in Table 5-3 were calculated using the adult mean pathway dose and risk
conversion factors (PDCFs and PRCFs) derived in Chapter 2. The distance from the source to the
receptor well was 528 ft.
Table 5-3: Radionuclide Source Term Limiting Concentrations -
Receptor Well at 528 ft
Nuclide
U-238
U-234
U-natural
Th-230
Ra-226
Ra-226+P
Pb-210
15 mrem/yr Dose Limit
(pCi/L)
97
93
95
25
1.2
2.4
97
(mg/L)
2.8E-01
1.5E-05
1.4E-01
1.2E-06
1.2E-09
3.1E-11
2.8E-01
10 4 LCF Lifetime Risk Limit
(pCi/L)
63
65
64
47
0.6
1.3
63
(mg/L)
1.8E-01
l.OE-05
9.4E-02
2.2E-06
6.3E-10
1.7E-11
1.8E-01
As Table 5-3 demonstrates, a lifetime risk limit of 10~4 LCF is slightly more restrictive than a
dose limit of 15 mrem/yr for all the radionuclides except Th-230. However, for all of the
radionuclides considered, dose- and risk-limiting concentrations are within a factor of two.
5-3
-------
Of the 31 unique excursion simulations (37 total) within the ore-bearing aquifer that met the
conductivity x gradient cutoff of 0.13 ft/day, the dose from uranium was <15mrem/yr in 13
simulations. The highest estimated dose from uranium at a receptor well 528 ft down-gradient
was 10,072 mrem/yr for Run 6i, while the lowest estimated dose was 1.69E-12 mrem/yr for
Run 5g. Table 5-4 summarizes these excursion simulations. Doses from uranium (U-234 +
U-238) are included in the last column for those simulations with a conductivity x gradient
product <0.13 ft/day. Runs where the uranium dose was less than 15 mrem/yr are highlighted in
yellow.
Table 5-4: Summary of Excursion Runs
Run
la
Ib
Ic
Id
le
If
lg
Ih
li
2a
2b
2c
2d
2e
2f
2g
2h
2i
3a
3b
3c
3d
3e
3f
3g
3h
3i
4a
4b
4c
4d
4e
4f
4g
4h
4i
5a
5b
5c
Hydraulic
Gradient
0.1
0.01
0.001
0.1
0.1
0.1
0.01
0.01
0.01
0.1
0.01
0.001
0.1
0.1
0.1
0.01
0.01
0.01
0.1
0.01
0.001
0.1
0.1
0.1
0.01
0.01
0.01
0.1
0.01
0.001
0.1
0.1
0.1
0.01
0.01
0.01
0.1
0.01
0.001
Hydraulic
Conductivity
(ft/day)
100
100
100
1
10
100
1
10
100
100
100
100
1
10
100
1
10
100
100
100
100
1
10
100
1
10
100
100
100
100
1
10
100
1
10
100
10
10
10
Well
Spacing
(ft)
250
250
250
250
250
250
250
250
250
50
50
50
50
50
50
50
50
50
250
250
250
250
250
250
250
250
250
50
50
50
50
50
50
50
50
50
150
150
150
Pumping
Array
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
7-spot
5 -spot
5 -spot
5 -spot
Pumping
Rate
(gpm)
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
153
51
51
51
Max. Relative
Concentration at
528ft
6.94E-03
6.54E-02
9.32E-02
2.14E-01
6.55E-02
6.94E-03
1.80E-04
9.27E-02
6.54E-02
9.62E-03
2.35E-02
8.75E-03
1.18E-02
2.35E-02
9.62E-03
3.39E-06
8.22E-03
2.35E-02
2.78E-03
3.53E-02
6.79E-02
1.41E-01
3.53E-02
2.78E-03
4.46E-05
6.75E-02
3.53E-02
1.25E-02
4.03E-02
9.01E-03
3.07E-02
4.03E-02
1.25E-02
2.14E-06
8.94E-03
4.03E-02
3.76E-02
6.79E-02
8.24E-04
Ore Zone
Thickness
(ft)
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
75
Unat
Dose3
(mrem/yr)
1.50E+03
3.44E+03
2.90E+00
1.49E+03
1.41E+02
1.90E+02
5.45E-02
1.32E+02
1.09E+03
2.27E+03
1.70E+00
1.09E+03
1.45E+02
4.94E+02
3.44E-02
1.44E+02
1.09E+03
1.33E+01
5-4
-------
Table 5-4: Summary of Excursion Runs
Run
5d
5e
5f
5g
5h
5i
6a
6b
6c
6d
6e
6f
6g
6h
6i
7a
7b
7c
7d
7e
7f
7g
7h
7i
Hydraulic
Gradient
0.01
0.01
0.01
0.001
0.001
0.001
0.1
0.01
0.001
0.01
0.01
0.01
0.001
0.001
0.001
0.1
0.01
0.001
0.01
0.01
0.01
0.001
0.001
0.001
Hydraulic
Conductivity
(ft/day)
1
10
100
1
10
100
10
10
10
1
10
100
1
10
100
10
10
10
1
10
100
1
10
100
Well
Spacing
(ft)
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
Pumping
Array
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
25 x 5 -spot
Pumping
Rate
fepm)
7.15
51
510
7.15
51
510
51
51
51
7.15
51
510
7.15
51
510
51
51
51
7.15
51
510
7.15
51
510
Max. Relative
Concentration at
528ft
1.53E-02
6.79E-02
1.01E-01
1.05E-16
8.24E-04
1.08E-01
3.31E-01
4.75E-01
1.30E-03
1.06E-01
4.75E-01
6.45E-01
3.07E-13
1.30E-03
6.26E-01
3.12E-01
6.61E-04
2.70E-04
8.11E-06
6.61E-04
1.10E-02
5.25E-05
2.70E-04
4.50E-04
Ore Zone
Thickness
(ft)
75
75
75
75
75
75
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Unat
Dose3
(mrem/yr)
2.46E+02
1.09E+03
1.69E-12
1.33E+01
1.74E+03
7.64E+03
2.09E+01
1.70E+03
7.64E+03
1.19E+00
2.09E+01
1.01E+04
1.06E+01
4.34E+00
3.78E-01
1.06E+01
8.44E-01
4.34E+00
7.24E+00
a - Dose at receptor well 528 ft down-gradient for simulations where hydraulic conductivity x gradient <0.13 ft/day
A breakdown of the unique simulations resulting in doses above and below 15 mrem/yr as a
function of the conductivity x gradient product is presented in Table 5-5. It is clear that for most
of the simulations with a dose of less than 15 mrem/yr, the product of hydraulic conductivity and
gradient is 0.01 or less. However, in 10% of the cases, a conductivity x gradient product of 0.01
resulted in doses greater than 15 mrem/yr.
Table 5-5: Number of Simulations Resulting in Various Dose Levels as Function of
Conductivity * Gradient Product
Conductivity (K) x Gradient (i)
Kxi = 0.1
KX 1 = 0.01
KX 1 = 0.001
Number of Simulations
Less Than 15 mrem/yr
2
7
3
Greater Than 15 mrem/yr
16
o
J
0
To evaluate the effects of retardation on ground water transport, distribution coefficients were
selected from the literature based on reasonable minimum values to reduce the travel time from
the source wellfield to a down-gradient receptor well. Minimum peak travel times to a receptor
well 528 ft down-gradient are about 20 years for uranium, about 3,000 years for radium and
about 15,000 years for thorium. The significance is that, if any lixiviant escapes undetected from
5-5
-------
the ring of monitor wells surrounding a wellfield, an appreciable amount of time will elapse
before the radionuclide arrives at the receptor well. The elapsed time will exceed the monitoring
times at any current ISL facilities.
Table 5-6 (which repeats Table 4-9) shows the maximum calculated dose for each radionuclide
from all excursion scenarios that were analyzed for dose and risk as shown in Table 5-4. These
values are mean adult exposures for a receptor well 528 ft down-gradient.
Table 5-6: Excursion Scenario Maximum Doses and Risks
Nuclide
Unat
Th-230
Ra-226+P
Dose (mrem/yr)
l.OE+04
2.4E+02
2.8E+04
Risk (LCF/yr)
1.4E-03
1.2E-05
4.8E-03
As described in Chapter 4 (Section 4.3.3), three surface leakage scenarios were evaluated:
(1) catastrophic spills ranging from 100,000 to 200,000 gallons, (2) a slow leak of 1 to 2 gpm for
period of 3 years, and (3) leaks varying from 1 to 40 gpm over a 28-day period. The highest
doses were incurred for scenarios involving a slow leak over a 3-year period, while the lowest
doses resulted from a 1-gpm surface leak over a 28-day period. In all cases, the mean annual
doses to an adult from U nat were greater than 15 mrem. Results for all leakage scenarios are
summarized in Table 5-7. These results emphasize the importance of detecting small leaks
through adequate instrumentation and frequent inspections of process piping. Automatic shut-off
controls should be used to minimize surface spills.
Table 5-7: Summary of Dose Rate for Leak Scenarios - Adult Male Exposed to U nat
at Receptor Well 328 ft Down-gradient
Leak Scenario
Surface Spill (100,000-200,000 gal.)
Slow Leak (1-2 gpm) - 3 Years
Variable Leak (1-40 gpm) - 28 days
Minimum Dose Rate (mrem/yr)
6.7E+01(RunL4)
9.5E+01 (Run LI 5)
3.2E+01 (RunL19)
Maximum Dose Rate (mrem/yr)
1.5E+02(RunL9)
1.7E+03 (RunL16)
8.7E+02(RunL24)
Doses and risks to non-standard receptors are compared to those for 90* percentile adults
(relative dose/risk = 1.00) in Table 5-8. The basis for comparison is the excursion scenario,
which occurs within the ore-bearing aquifer and where the receptor obtains water from a well
528 ft down-gradient. Doses and risks for the Mean Infant and Native American were greater
than for the 90* percentile adult receptor for all radionuclides evaluated. Relative doses for the
mean teenager and the mean child (except for Ra-226+P) were lower than for the 90* percentile
adult. For example, doses and risks to Native Americans were greater by factors of about 1.2 to
1.8 than for the 90th percentile adult. Clearly, doses to several of the non-standard receptors do
not fall within a reasonable upper limit for the standard receptor doses.
5-6
-------
Table 5-8: Comparison of Excursion Scenario Non-standard Receptor Doses and
Risks Relative to 90th Percentile Adult
Receptor
90th Percentile Adult
Mean Teenager
Mean Child
Mean Infant
Native American
Nuclide
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Unat
Th-230
Ra-226+P
Maximum
Relative Dose
1.00
1.00
1.00
0.55
0.33
0.89
0.50
0.27
1.17
1.55
3.92
3.17
1.44
1.78
1.20
Relative Risk
1.00
1.00
1.00
1.18
0.81
0.99
2.06
0.84
1.88
3.46
2.34
1.97
1.74
1.76
1.26
5-7
-------
6.0 REFERENCES
Abbott M.L., and A.S. Rood, 1993. "COMIDA: A Radionuclide Food Chain Model for Acute
Fallout Deposition," Idaho National Engineering Laboratory, EGG-GEO-1O367, Revision O.
November 1993.
ANL (Argonne National Laboratory) 2000. "Development of Probabilistic RESRAD 6.0 and
RESRAD-BUILD 3.0 Computer Codes," prepared for U.S. Nuclear Regulatory Commission,
NUREG/CR-6697, ANL/EAD/TM-98, November.
ANL (Argonne National Laboratory) 2001. "User's Manual for RESRAD Version 6,"
Environmental Assessment Division, ANL/EAD-4, July.
ATSDR (Agency for Toxic Substances and Disease Registry) 2010. "Public Health Assessment
for Midnite Mine Site, Wellpinit, Stevens County, Washington, EPA Facility ID:
WAD980978753, May 19.
CBR (Crow Butte Resources, Inc.). 2007. "SUA-1535 License Renewal Application."
Crawford, Nebraska: Crow Butte Resources, Inc. November 2007.
CDC (Centers for Disease Control and Prevention) 2011. "Breastfeeding Report Card — United
States, 2011" August 2011.
Chowdhury, A.H., and R.E. Mace, 2007. Groundwater resource evaluation and availability
model of the Gulf Coast aquifer in the Lower Rio Grande Valley of Texas: Texas Water
Development Board, Report 368, 119 p.
CNWRA (Center for Nuclear Waste Regulatory Analyses) 2001. "A Baseline Risk-Informed,
Performance-Based Approach for In Situ Leach Uranium Extraction Licensees." NUREG-
CR/6733. September 2001.
COGEMA2008. WellfieldRestoration Report Christensen Ranch Project, Wyoming. COGEMA
Mining, Inc, and Petrotek Engineering Corp, NRC ADAMS Accession Number: ML081060131.
March 5, 2008.
Crow Butte 1995. Application for Renewal of USNRC Source Materials License SUA-1534.
Crow Butte Resources, Inc. ML082140217.
Davis, J.A. and G.P. Curtis, 2007. NUREG/CR-6870, "Consideration of Geochemical Issues in
Groundwater Restoration at Uranium In-Situ Leaching Mining Facilities." Washington, DC:
Nuclear Regulatory Commission. January 2007.
DBS&A (Daniel B. Stephens and Associates, Inc.) 2007. "Evaluation of Potential Impacts
Related to Proposed Uranium Mining in Goliad County, Texas." Prepared for Goliad County
Groundwater Conservation District. June 25, 2007.
6-1
-------
Deutsch W.J., WJ. Martin, L.E. Eary, and RJ. Serne, 1985. Methods of Minimizing Ground-
Water Contamination from In Situ Leach Uranium Mining: Final Report. PNL-5319, U.S.
Nuclear Regulatory Commission (Ed.), Washington, DC.
DOE (U.S. Department of Energy) 1997. "Screening Assessment and Requirements for a
Comprehensive Assessment," DOE/RL-96-16, Revision 0, April.
DOE (U.S. Department of Energy) 2010. "River Corridor Baseline Risk Assessment," DOE/RL-
2007-21, Volume II, "Human Health Risk Assessment," Part 1, Draft C, December.
Eckerman, K.F., R.W. Leggett, C.B. Nelson, J.S. Puskin, and A.C.B. Richardson, 1999. "Federal
Guidance Report No. 13: Cancer Risk Coefficients for Environmental Exposure to
Radionuclides," EPA-402-R-99-001, prepared for U.S. Environmental Protection Agency; Office
of Radiation and Indoor Air, September.
EPA (U.S. Environmental Protection Agency) 1990. "National Contingency Plan," Federal
Register, Volume 55, pages 8665-8865. March 8, 1990.
EPA (U.S. Environmental Protection Agency) 1991. "National Primary Drinking Water
Regulation; Final Rule," Federal Register 56(20):3526-3597, January 30.
EIA (Energy Information Administration) 1995. DOE/EIA-0592, "Decommissioning of U.S.
Uranium Production Facilities." Energy Information Administration, Office of Coal, Nuclear,
Electric, and Alternate Fuels, Washington, DC. February 1995.
EPA (U.S. Environmental Protection Agency) 1999a. "Estimating Radiogenic Cancer Risks
Addendum: Uncertainty Analysis," Office of Radiation and Indoor Air, EPA 402-R-99-003. May
1999.
EPA (U.S. Environmental Protection Agency) 1999b. "Understanding Variation in Partition
Coefficient, Kd, Values: Volume II: Review of Geochemistry and Available Kd Values for
Cadmium, Cesium, Chromium, Lead, Plutonium, Radon, Strontium, Thorium, Tritium (3H) and
Uranium," EPA 402-R-99-004B. Office of Radiation and Indoor Air, Office of Solid Waste and
Emergency Response, EPA [and] Office of Environmental Restoration, U.S. Department of
Energy, Washington, DC. August 1999.
EPA (U.S. Environmental Protection Agency) 2001. "Public Health and Environmental
Radiation Protection Standards for Yucca Mountain, NV: Background Information Document
for 40 CFR Part 197," EPA 402-R-01-004. June 2001.
EPA (U.S. Environmental Protection Agency) 2002. Report No. 13, Cancer Risk Coefficients for
Environmental Exposure to Radionuclides: Updates and Supplements.
EPA 2004. "Understanding Variation in Partition Coefficient, Kd, Values: Volume III: Review of
Geochemistry and Available Kd Values for Americium, Arsenic, Curium, Iodine, Neptunium,
Radium and Technetium. " EPA 402-R-99-004C. Office of Radiation and Indoor Air. July 2004.
6-2
-------
EPA 2005. "Guidance on Selecting Age Groups for Monitoring and Assessing Childhood
Exposures to Environmental Contaminants," EPA/630/P-03/003F, November 2005.
EPA (U.S. Environmental Protection Agency) 2008. "Radiation Protection Standards for Yucca
Mountain," Federal Register, Volume 73, pages 61256- 61289. October 15, 2008.
EPA (U.S. Environmental Protection Agency) 2009. "2009 Water Sense Single-Family New
Home Specification, Supporting Statement," Version 1.0, December 9.
EPA (U.S. Environmental Protection Agency) 201 la. "2011 Edition of the Drinking Water
Standards and Health Advisories," Office of Water, EPA 820-R-l 1-002. Januray 2011.
EPA (U.S. Environmental Protection Agency) 201 Ib. "Exposure Factors Handbook: 2011
Edition," National Center for Environmental Assessment, EPA/600/R-090/052F. September.
Freeze, R.A. and J. Cherry, 1979. "Groundwater," Prentice-Hall, Englewood Cliffs, New Jersey.
Gelhar, L.W., C. Welty, and K.R. Rehfeldt, 1992. "A critical review of data on field-scale
dispersion in aquifers." Water Resources Research 28(7): 1955-1974.
Godbee, H.W., 1973. "Use of Evaporation for the Treatment of Liquids in the Nuclear Industry,"
ORNL-4790. Oak Ridge National Laboratory, Oak Ridge, Tennessee. September 1973.
Harbaugh, A.W., E.R. Banta, M.C. Hill, and M.G. McDonald, 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.
Harper, B.L., A.K. Harding, T. Waterhous, and S.G. Harris, 2007. "Traditional Tribal
Subsistence Exposure Scenario and Risk Assessment Guidance Manual," Oregon State
University, Department of Public Health and Confederated Tribes of the Umatilla Indian
Reservation, August.
Harper, B., and D. Ranco, 2009. "Wabanaki Traditional Cultural Lifeways Exposure Scenario,"
Associated Environmental Scientists & Engineers (AESE) and Dartmouth College, prepared for
the U.S. Environmental Protection Agency in collaboration with the Maine Tribes, July 9.
Harris, S.G. and B.L. Harper, 2004. "Exposure Scenario for CTUIR Traditional Subsistence
Lifeways," Department of Science & Engineering, Confederated Tribes of the Umatilla Indian
Reservation, September 15.
Hem, J.D., 1985. "Study and Interpretation of the Chemical Characteristics of Natural Water."
USGS Water Supply Paper 2254. Third edition. Reston, Virginia: U.S. Geological Survey.
Hopke, P.K., 2006. "Exposure and Risks from Showering with Radon-Rich Water." Presented at
"Workshop on Radon Occurrence, Health Risks and Policy, With an Emphasis on Radon in
Ground Water Drinking Supplies" (October 4).
6-3
-------
Hoy, R.N., 2006. Baseline Ground Water Quality Conditions at In Situ Uranium Wellfields in
Wyoming. Paper presented at the 2006 Billings Land Reclamation Symposium, June 4-8, 2006.
Billings Montana. http://www.asmr.us/Publications/Conference%20Proceedings/
2006%20Billings/0246-Hov-WY-1 .pdf
HRI1993. "Church Rock Uranium Project Revised Environmental Report. March 16, 1993.
ICRP (International Commission on Radiological Protection), 1996. "Age-dependent Doses to
the Members of the Public from Intake of Radionuclides Part 5, Compilation of Ingestion and
Inhalation Coefficients," Publication 72, Annals of the ICRP Volume 26/1, September.
ICRP (International Commission on Radiological Protection), 2001. "Doses to the Embryo and
Fetus from Intakes of Radionuclides by the Mother," Publication 88, Annals of the ICRP
Volume 31/1-3, December.
ICRP (International Commission on Radiological Protection), 2004. "Doses to Infants from
Ingestion of Radionuclides in Mothers' Milk," Publication 95, Annals of the ICRP Volume 34/-4,
September.
ICRP (International Commission on Radiological Protection), 2006. "Assessing Dose of the
Representative Person for the Purpose of the Radiation Protection of the Public," Publication
101, Annals of the ICRP Volume 36/3, September.
ISCORS (Interagency Steering Committee on Radiation Standards), 2002. "A Method for
Estimating Radiation Risk from Total Effective Dose Equivalent (TEDE), Final Report,"
ISCORS Technical Report 2002-02.
Lalman, D., 2004. "Nutrient Requirements of Beef Cattle," E-974, Oklahoma State University,
Oklahoma Cooperative Extension Service.
Looper, M.L., and D.N. Waldner, 2002. "Water for Dairy Cattle," New Mexico State University
and Oklahoma State University, Guide D-107, February.
Lost Creek 2007. Lost Creek Project Technical Report, Application for U.S. NRC Source
Material License," Docket 40-9068, Volume 2. ML081060505.
Mackin, P.C., D. Daruwalla, J. Winterle, M. Smith, and D.A. Pickett, 2001. "A Baseline Risk-
Informed Performance-Based Approach for In-Situ Leach Uranium Extraction Licensees."
NUREG/CR-6733, U.S. Nuclear Regulatory Commission, Washington, DC. September 2001.
Napier BA, Peloquin, Strenge DL, Ramsdell JV Jr., 1988. "GENII - the Hanford Environmental
Radiation Dosimetry Software System. Volume 1: Conceptual Representation, Volume 2: Users'
Manual, Volume 3: Code Maintenance Manual," PNL-6584, Pacific Northwest Laboratory,
Richland, Washington.
6-4
-------
NAS (National Academy of Sciences) 1995. "Technical Bases for Yucca Mountain Standards,"
National Research Council, Committee on Technical Bases for Yucca Mountain Standards,
National Academy Press, Washington, DC. 1995.
NAS (National Academy of Sciences) 1999a, "Risk Assessment of Radon in Drinking Water,"
Commission on Life Sciences, Washington, DC.
NAS (National Academy of Sciences) 1999b, "Health Effects of Exposure to Radon, BEIR VI,"
Committee on Health Risks of Exposure to Radon, Washington, DC.
NAS (National Academy of Sciences) 2000. "Nutrient Requirements of Beef Cattle: Seventh
Revised Edition: Update 2000," National Research Council.
NAS (National Academy of Sciences) 2001. "Nutrient Requirements of Dairy Cattle: Seventh
Revised Edition," National Research Council.
NCRP (National Council on Radiation Protection and Measurements) 1999. "Recommended
Screening Limits for Contaminated Surface Soil and Review of Factors Relevant to Site-Specific
Studies," Report No. 129, January 29.
Nicot, J-P, B.R. Scanlon, C. Yang, and J.B. Gates, 2010. "Geological and Geographical
Attributes of the South Texas Uranium Province." A report prepared for the Texas Commission
on Environmental Quality. April 2010.
NRC (U.S. Nuclear Regulatory Commission) 1977. "Calculation of Annual Doses to Man from
Routine Releases of Reactor Effluents for the Purpose of Evaluating Compliance With 10 CFR
Part 50, Appendix I," Regulatory Guide 1.109, Revision 1, October.
NRC (U.S. Nuclear Regulatory Commission) 1978. "Environmental Statement Related to
Operation of Wyoming Mineral Corporation Irigaray Solution Mining Project." NUREG-0399.
Docket No. 40-8502. ML091120676.
NRC (U.S. Nuclear Regulatory Commission) 1985. "Calculation of Releases of Radioactive
Materials in Gaseous and Liquid Effluents from Pressurized Water Reactors," NUREG-0017,
Revision 1, April.
NRC (U.S. Nuclear Regulatory Commission) 1989. "Environmental Assessment by the Uranium
Recovery Field Office in Consideration of an Application for a Source Material License for
Ferret Exploration Company of Nebraska, Crow Butte Commercial In situ Leach Facility,
Dawes County, Nebraska." Docket No, 40-8943. December 12, 1989. ML080740160.
NRC (U.S. Nuclear Regulatory Commission) 1991. "A Regulatory Analysis on Emergency
Preparedness for Fuel Cycle and Other Radioactive Material Licensees," NUREG-1140,
Reprinted August.
NRC (U.S. Nuclear Regulatory Commission) 1997. "Final Environmental Impact Statement to
Construct and Operate the Crownpoint Uranium Solution Mining Project, Crownpoint, New
6-5
-------
Mexico." NUREG-1508. U.S. Nuclear Regulatory Commission, Washington, DC. February
1997.
NRC (U.S. Nuclear Regulatory Commission) 1998. "Effects on Radionuclide Concentrations by
Cement/Ground-Water- Interactions in Support of Performance Assessment of Low-Level
Radioactive Waste Disposal Facilities." NUREG/CR-6377. Prepared by K.M. Krupa and RJ.
Serne. Pacific Northwest National Laboratory, Richland, Washington.
NRC (U.S. Nuclear Regulatory Commission) 2000. "Inspection Report 40-8964/00-01- Smith
Ranch." U.S. Nuclear Regulatory Commission: Washington, DC. ML003683025.
NRC (U.S. Nuclear Regulatory Commission) 2003. "Standard Review Plan for In Situ Leach
Uranium Extraction License Applications." Final Report. NUREG-1569. Prepared by J. Lusher.
Washington, DC. June 2003. ML032250177.
NRC (U.S. Nuclear Regulatory Commission) 2006. "Environmental Assessment for the Addition
of the Reynolds Ranch Mining Area to Power Resources, Inc.'s Smith Ranch/Highlands
Uranium Project Converse County, Wyoming." Source Material License No. SUA-1548.
Docket No. 40-8964. Nuclear Regulatory Commission, Washington, DC. 2006.
NRC (U.S. Nuclear Regulatory Commission) 2007. "Environmental Assessment Construction
and Operation of In Situ Leach Satellite SR-2 Amendment No. 12 to Source Material License
No. SUA-1548." Power Resources, Inc., Smith Ranch-Highland Uranium Project (SR-HUP),
Converse County Wyoming. ML071500581.
NRC (U.S. Nuclear Regulatory Commission) 2008. "Compliance with 10 CFR 40.42's Timely
Decommissioning Requirements." Letter (July 7) from K.I. McConnell to S. Collings, Power
Resources, Inc., Nuclear Regulatory Commission, Washington, DC. 2008.
NRC (U.S. Nuclear Regulatory Commission) 2009a. "Uranium Recovery Policy Regarding:
(1) The Process for Scheduling Licensing Reviews of Applications for New Uranium Recovery
Facilities and (2) The Restoration of Groundwater at Licensed Uranium In Situ Recovery
Facilities." Regulatory Information Summary 2009-05. ADAMS Accession No. ML083510622.
NRC. April 29, 2009a.
NRC (U.S. Nuclear Regulatory Commission) 2009b. "Generic Environmental Impact Statement
for In-Situ Leach Uranium Milling Facilities." Chapters 1 through 4, Final Report. Nuclear
Regulatory Commission, Washington, DC. May 2009b.
NRC (U.S. Nuclear Regulatory Commission) 2009c. "Environmental Impact Statement for the
Nichols Ranch ISR Project in Campbell and Johnson Counties Wyoming." NUREG-1910
Supplement 2. Draft Report for Comment. ML093340536.
Pawel D.J., R.W. Leggett, K.F. Eckerman, and C.B. Nelson, 2007. "Uncertainties in Cancer Risk
Coefficients for Environmental Exposure to Radionuclides, An Uncertainty Analysis for Risk
Coefficients Reported in Federal Guidance Report No. 13," U.S. Environmental Protection
Agency and Oak Ridge National Laboratory, ORNL/TM-2006/583, January.
6-6
-------
Pelizza, M.S., 2008. "In-Situ Recovery of Uranium." Southwest Hydrology 7:28-29.
Peterson, H.T., Jr, 1983. "Terrestrial and Aquatic Food Chain Pathways," in Till JE, Meyer HR.
Radiological Assessment, A Textbook on Environmental Dose Analysis, NUREG/CR-3332,
U.S. Nuclear Regulatory Commission, Washington, DC.
PNL (Pacific Northwest Laboratory) 1988. "GENII - The Hanford Environmental Radiation
Dosimetry Software System, Volume 1: Conceptual Representation," PNL-6584, December.
PNNL (Pacific Northwest National Laboratory) 2006. "Alternative Conceptual Models for
Assessing Food-Chain Pathways in Biosphere Models," prepared for U.S. Nuclear Regulatory
Commission, Office of Nuclear Regulatory Research, NUREG/CR-6910, PNNL-15872, June.
PNNL (Pacific Northwest National Laboratory) 2010. "Evaluation of Soil Flushing for
Application of Deep Vadose Zone in the Hanford Central Plateau." PNNL-1993 8. MJ. Truex
et al., Pacific Northwest National Laboratory.
Power Resources 2004. Mine Unit B Ground Water Restoration Report, Smith Ranch - Highland
Uranium Project. NRC ADAMS Accession Number: ML091831100.
PRI (Power Resources Inc.) 2006. "North Butte ISL Satellite Project, Campbell County,
Wyoming." Volume 1. ML061740064.
Rice, G., 2006. "Effects of URI's Kingsville Dome Mine on Groundwater Quality." Final
Report, prepared for the Kleberg County URI Citizen Review Board.
Rittmann, P.O., 1999. "Exposure Scenarios and Unit Factors for Hanford Tank Waste
Performance Assessments," Fluor Government Group, HNF-SD-WM-TI-707, Revision 1,
December.
Rittmann, P.D., 2004. "Exposure Scenarios and Unit Factors for Hanford Tank Waste
Performance Assessments," Fluor Government Group, HNF-SD-WM-TI-707, Revision 4, May.
Rumbaugh, J.O., and D.B. Rumbaugh, 1998. "Groundwater Vistas." Version 5.16.
Environmental Simulations.
SC&A 2008. "Database Summary Analysis Report: Uranium Mills and In-Situ Leach
Facilities." Contract EP-D-05-002, Work Assignment 4-17, Task 6. SC&A, Inc., Vienna,
Virginia. September 2008.
SC&A 201 la. "Considerations Related to Post-Closure Monitoring of Uranium In-Situ Leach/In-
Situ Recovery (ISL/ISR) Sites." Contract Number EP-D-10-042, Work Assignment 2-06, Task
5. SC&A, Inc., Vienna, Virginia.
SC&A 201 Ib. "Modeling of Heap Leach and In-Situ Leaching Operations." Contract Number
EP-D-10-042, Work Assignment 1-11, Task 4. SC&A, Inc., Vienna, Virginia. March 26, 2011.
6-7
-------
SC&A 201 Ic. "Modeling of Heap Leach and In-Situ Leaching Operations." Contract Number
EP-D-10-042, Work Assignment 2-06, Task 2. SC&A, Inc., Vienna, Virginia. May 4, 2011.
Schramke, J.A., H. Demuth, and M.S. Pelizza, 2009. "Geochemistry of Uranium In-Situ Leach
Aquifers After Restoration." Presented at U2009 Global Uranium Symposium, Keystone,
Colorado, May 2009.
SNL (Sandia National Laboratory) 1999. "Residual Radioactive Contamination From
Decommissioning, Parameter Analysis, Draft Report for Comment" prepared for U.S. Nuclear
Regulatory Commission, NUREG/CR-5512, Vol. 3, SAND99-2148, October.
Terry, Robert 2011. Personnel communication, U.S. Environmental Protection Agency,
Region 9, October.
Trinity (Trinity Engineering Associates, Inc.) 2007. "CAP88-PC Version 3.0 User Guide,"
prepared for the U.S. Environmental Protection Agency, December 9.
USCB (U.S. Census Bureau) 2011. "Profile of General Population and Housing Characteristics:
2010," DP-1.
USDA (U.S. Department of Agriculture) 2004. "2002 Census of Agriculture, Farm and Ranch
Irrigation Survey (2003)," AC-02-SS-1, November.
USDA (U.S. Department of Agriculture) 2009. "2007 Census of Agriculture, Summary and State
Data, Volume 1, Geographic Area Series, Part 51," AC-07-A-51, updated December.
USDA (U.S. Department of Agriculture) 2011. "2010 Vegetable Summary," National
Agricultural Statistics Service, January.
UT (University of Texas) 2010. "Geological and Geographical Attributes of the South Texas
Uranium Province." Jackson School of Geosciences, University of Texas at Austin.
Whicker, F.W. and T.B. Kirshner, 1987. "PATHWAY: A Dynamic Food-Chain Model to
Predict Radionuclide Ingestion After Fallout Deposition." Health Physics, 52(6):717-737.
WSDOH (Washington State, Department of Health) 2003. "Radiological Risk Assessment, Low-
Level Radioactive Waste Disposal Site, Richland, WA," Andrew H. Thatcher, October.
Yu C., A.J. Zielen, J.J. Cheng, D.J. LePoire, E. Gnanapragasam, S. Kamboj, J. Arnish, A. Wallo
III, W.A. Williams, H. Peterson, 2002. "User's Manual for RESRAD Version 6," ANL/EAD-4,
Argonne National Laboratory, Argonne, Illinois.
Zheng, C., and P. Patrick Wang, 1999. "MT3DMS: A modular three-dimensional multispecies
transport model for simulation of advection, dispersion, and chemical reactions of contaminants
in groundwater systems; documentation and user's guide," Contract Report SERDP-99-1, U.S.
Army Engineer Research and Development Center, Vicksburg, Mississippi.
6-8
------- |