Radionuclide dose assessment of groundwater migrating from ...

UNLV Retrospective Theses & Dissertations

1-1-1998

Radionuclide dose assessment of groundwater migrating from Radionuclide dose assessment of groundwater migrating from

the Benham Events, Nts: A solute flux approach the Benham Events, Nts: A solute flux approach

Lloyd T Desotell University of Nevada, Las Vegas

Follow this and additional works at: https://digitalscholarship.unlv.edu/rtds

Repository Citation Repository Citation Desotell, Lloyd T, "Radionuclide dose assessment of groundwater migrating from the Benham Events, Nts: A solute flux approach" (1998). UNLV Retrospective Theses & Dissertations. 914. http://dx.doi.org/10.25669/5d7d-wudy

This Thesis is protected by copyright and/or related rights. It has been brought to you by Digital Scholarship@UNLV with permission from the rights-holder(s). You are free to use this Thesis in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/or on the work itself. This Thesis has been accepted for inclusion in UNLV Retrospective Theses & Dissertations by an authorized administrator of Digital Scholarship@UNLV. For more information, please contact [email protected].

http://library.unlv.edu/

http://library.unlv.edu/

https://digitalscholarship.unlv.edu/rtds

https://digitalscholarship.unlv.edu/rtds?utm_source=digitalscholarship.unlv.edu%2Frtds%2F914&utm_medium=PDF&utm_campaign=PDFCoverPages

http://dx.doi.org/10.25669/5d7d-wudy

mailto:[email protected]

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI

films the text directly from the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter 6ce, while others may be

from any type of computer printer.

The quality of this reproduction Is dependent upon the quality of the

copy submitted. Broken or indistinct print, colored or poor quality

illustrations and photographs, print bleedthrough, substandard margins,

and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete

manuscript and there are missing pages, these will be noted. Also, if

unauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand comer and

continuing from left to right in equal sections with small overlaps. Each

original is also photographed in one exposure and is included in reduced

form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6” x 9” black and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact UMI directly to

order.

UMIA Bell & Hbwelllnfixmation Company

300 North Zeeb Road, Ann Aibor MI 48106-1346 USA 313/7614700 800/321-0600

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

RADIONUCLIDE DOSE ASSESSMENT OF GROUNDWATER MIGRATING

FROM THE BENHAM EVENT. NTS; A SOLUTE FLUX APPROACH

by

Lloyd T. Desotell

Bachelor o f Arts University o f Nevada, Las Vegas

1993

A thesis submitted in partial satisfaction o f the requirements for degree o f

Master of Science

in

Water Resources Management

Department of Geoscience University of Nevada, Las Vegas

August 1998


UMI Number: 1392318

Copyright 1999 by Desotell, Lloyd T.

All rights reserved.

UMI Microform 1392318 Copyright 1998, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI300 North Zeeb Road Ann Arbor, MI 48103


I7NTV Thesis ApprovalThe Graduate College University of Nevada, Las Vegas

July 25. 1998

The Thesis prepared by

Lloyd I . Desotell

Entitled

RADIONUCLIDE DOSE ASSESSMENT OF GROUNDWATER MIGRATING FROM THE BENHAM

EVENT. NTS; A SOLUTE FLUX APPROACH____________________________________

is approved in partial fulfillment of the requirements for the degree of

Master o f Science in Water Resources Management

Examination Committee Member

ExaminationAIommittee

Graduate College Faculty Representative

Examination Committee Chair

Dean of the Ctful uite College

U


ABSTRACT

Radionuclide Dose Assessment of Groundwater Migrating from the Benham Event, NTS; a Solute Flux Approach

By

Lloyd T. Desotell

Dr. Roko Andricevic, Examination Committee Chair Professor o f Hydrology

University of Nevada, Las Vegas

Radionuclide Dose Assessment o f Groundwater Migrating from the Benham Event, NTS;

a Solute Flux Approach presents results o f a preliminary dose-based modeling study o f

radionuclide transport from the Benham subsurface nuclear detonation at the Nevada Test

Site. This investigation utilized a radiologic source term generated from publicly

available data and consists o f forty-two radionuclides that may be present in sufficient

quantities to pose a potential human health risk at the source. To avoid the complex near

cavity chemistry (e.g. solubility o f actinides), conservative assumptions were used to

estimate a hydrologie source term. By using a two-dimensional stochastic groundwater

contaminant transport code, radionuclide transport was modeled from the Benham event

to control planes 10, 20, and 28 km downgradient assuming a pulse injection. The 95%

confidence level o f peak flux-averaged concentration at the control plane was compared

to species specific concentration thresholds that would produce a 4 mrem/yr committed

iii


effective dose equivalent assuming 2L/day groundwater ingestion. Species with

predicted concentrations above threshold values were modeled to more distant

boundaries, while species w ith predictions that fe ll below their respective threshold were

eliminated from further consideration. Twenty o f the in itia l forty-two radionuclides have

predicted 95% level concentrations above threshold concentrations at Oasis Valley.

Thirteen o f these twenty are actinides. Actinide results are believed to be unreasonably

high due to assumptions made regarding solubility and potential for colloidal transport.

i29i i4ç and *̂C1 were identified as the radionuclides that have the most potential to

exceed concentration thresholds at the assumed discharge location o f Oasis Valley.

IV


ACKNOWLEDGMENTS

First I would like to thank Tod Johnson for his direction and encouragement

which made this project possible. I would also like to acknowledge the members o f my

graduate committee, Dr. Roko Andricevic (Desert Research Institute), Dr. David

Kreamer (Water Resources Management), Dr. Claus Stetzenbach (Harry Reid Center for

Environmental Studies) and Dr. Mark Rudin (Health Physics) for their assistance.


TABLE OF CONTENTS

ABSTRACT....................................................................................................................... iii

ACKNOW LEDGMENTS.................................................................................................. v

LIST OF FIGURES......................................................................................................... v iii

LIST OF TABLES..............................................................................................................ix

CHAPTER I INTRO DUCTIO N........................................................................................1Thesis O b jective ................................................................................................................2

CHAPTER 2 TH EO R Y......................................................................................................3

CHAPTER 3 METHODOLOGY....................................................................................... 5Lagrangian Transport Form ulation........................................................................ 7Kinematical Relationships..................................................................................... 10

CHAPTER 4 SOLUTE MASS F LU X ..............................................................................11Mean and V ariance............................................................................................... 12Effect o f Sorption and Decay..................................................................................14Conversion o f Solute Flux to the Flux-Averaged Concentration.......................... 15

CHAPTER 5 CASE STUDY.............................................................................................17Hydrogeologic Setting............................................................................................21Source Term ...........................................................................................................22Threshold Values................................................................................................... 24In itia l Distribution o f Radionuclides......................................................................25Solubility o f Radionuclides....................................................................................27Spatial Variability in Hydraulic Conductivity....................................................... 30Source Size............................................................................................................. 31Sampling Detection W indow ................................................................................32Correlation Scale................................................................................................... 32Aquifer Thickness................................................................................................. 33Radionuclide Retardation.....................................................................................33Colloid-Facilitated Transport................................................................................. 37Selection o f Control Planes....................................................................................38Mean Groundwater Velocity.................................................................................. 38

vi


Procedure................................................................................................................ 39Conversion o f Flux-Averaged Concentration to Committed EffectiveDose Equivalent...................................................................................................... 39

CHAPTER 6 RESULTS.................................................................................................... 41Sensitivity Analysis.......................................................................................... 49Uncertainty in Mean Plume Position............................................................... 57

CHAPTER 7 DISCUSSION.............................................................................................58Evaluation o f Method Selection............................................................................. 60

CHAPTER 8 CONCLUSION.......................................................................................... 61

REFERENCES...................................................................................................................62

V ITA .................................................................................................................................... 66

vu


LIST OF FIGURES

Figure 1 Plume schematic................................................................................................4Figure 2 Absolute Dispersion and Relative D ispersion................................................. 8Figure 3 Definition sketch o f particle displacement........................................................ 9Figure 4 State o f Nevada map.........................................................................................18Figure 5 Location o f detonations near or below the water table................................... 19Figure 6 Location o f Benham event...............................................................................20Figure 7 Breakthrough o f 95% Level o f Committed Effective Dose Equivalent

at Oasis V a lle y ...............................................................................................48Figure 8 Breakthrough o f 95% Level o f Committed Effective Dose Equivalent

o f Conservative Species at Oasis V alley........................................................48Figure 9 Summation o f Committed Effective Dose Equivalent at Oasis Valley. . . . 50Figure 10 *“’ l Committed Effective Dose Equivalent at Oasis V a lley........................... 50Figure 11 Sensitivity o f '"’ l Concentrations at Oasis Valley to Source Size................. 52Figure 12 Sensitivity o f Concentrations at Oasis Valley to Source S iz e ................. 52Figure 13 Sensitivity o f '̂ °Eu Concentrations at Oasis Valley to Mean

Groundwater Velocity..................................................................................... 54Figure 14 Sensitivity o f ‘"’ l Concentrations at Oasis Valley to Mean

Groundwater Velocity..................................................................................... 54Figure 15 Sensitivity o f Concentrations at Oasis Valley to Variance

in Hydraulic Conductivity..............................................................................55Figure 16 Sensitivity o f ‘"’ l Concentrations at Oasis Valley to Variance

in Hydraulic Conductivity..............................................................................55Figure 17 Sensitivity o f ̂ ’ Pu Concentrations at Oasis Valley to Retardation

Factor.............................................................................................................. 56Figure 18 Sensitivity o f ̂ S r Concentrations at Oasis Valley to Retardation

Factor............................................................................................................. 56

vui


LIST OF TABLES

Table 1Table 2Table 3Table 4Table 5Table 6Table 7Table 8Table 9Table 10Table 11Table 12Table 13Table 14Table 15Table 16Table 17Table 18Table 19Table 20

Physical Setting o f Benham S ite ...................................................................... 21Radiologic Source T e rm ...................................................................................26Assumptions for Near Cavity Distribution o f Radionuclides.......................... 27Solubilities o f Selected Radionuclides.............................................................28W ell J-13 Water Chemistry Data...................................................................... 29Retardation Factors for Western S o il...............................................................35Median Retardation Factors from a Variety o f Yucca Mountain Tuffs........... 36Assumed Retardation Factors........................................................................... 36Collodial Transport Parameters for Frenchman Flat........................................ 38Parameter Summary..........................................................................................40Fission Products Above Threshold Values at 10 k m ...................................... 43Activation Products Above Threshold Values at 10 km.................................. 43Actinides Above Threshold Values at 10 km ...................................................44Fission Products Above Threshold Values at 20 k m ...................................... 44Activation Products Above Threshold Values at 20 km ................................. 45Actinides Above Threshold Values at 20 km ...................................................45Fission Products Above Threshold Values at 28 k m ...................................... 46Activation Products Above Threshold Values at 28 km ................................. 46Actinides Above Threshold Values at 28 km ...................................................47Maximum Observed Actinide Concentrations................................................ 47

IX


CHAPTER 1

INTRODUCTION

W ith the escalating reliance on groundwater as a potable water supply, subsurface

environmental contamination has become a topic o f increasing interest. Groundwaters

may become contaminated from a variety o f activities including the disposal o f wastes,

use o f fertilizers or the storage o f fuels. The magnitude o f any groundwater quality

problem depends on factors such as the spatial extent o f contamination, the amount and

physical and chemical characteristics o f the contaminants involved, characteristics o f the

subsurface and the potential effect on groundwater usage. O f particular concern to

southern Nevadans is the migration o f radionuclides from high level waste storage or

subsurface nuclear testing, due to its potential to pose a long-term threat to aquifer

quality. Aquifer quality and health effects resulting from ingestion o f radionuclide

contaminated groundwater are directly related to contaminant concentrations.

Accordingly, accurate prediction o f concentration is critical in making prudent risk

assessment and remedial decisions.


Thesis Objectives

The objectives o f this thesis are to:

• Model the transport o f radionuclides from a single subsurface nuclear detonation

using a stochastic contaminant transport code.

• Determine which radionuclides are potentially significant with respect to health-based

criteria at multiple compliance boundaries.

• Use conservative assumptions to estimate the time to first potential exposure in

excess o f regulatory thresholds at Oasis Valley.

• Identify which radionuclides and hydrologie transport parameters need further study.

• Evaluate the method used at a single site with the intent o f applying it to other

locations.


CHAPTER 2

THEORY

It is assumed that groundwater flow is incompressible and steady through a

heterogeneous aquifer o f spatially variable hydraulic conductivity K{x). Groundwater

seepage velocity vector V{x) satisfies the continuity equation, V • V = 0, and is related to

the hydraulic conductivity and hydraulic head h through Darcy’s law V = - i ^ n ) Vh,

where n is the effective porosity. Additionally, it is assumed that although the steady

heterogeneous flow implies irregular streamlines, a mean groundwater flow direction can

be identified (figure 1). Assuming at t = to, a solute is instantaneously released into the

flow field over a finite volume, a solute plume is formed and advected downgradient by

the flow field, towards a control plane (figure 1 ) for t > to. Due to velocity fluctuations

caused by scales o f heterogeneity smaller than the plume size, the plume is distorted in an

irregular manner. Velocity fluctuations caused by scales o f heterogeneity which are

larger than the scale o f the plume size w ill result in the plume meandering relative to the

mean direction o f flow. The mean and covariance function o f V are considered known,

having been derived from the continuity equation and Darcy’s law using a first-order

approximation (e.g. Rubin and Dagan, 1992).


MEANFLOW

SOURCE AREA A

ACCESSIBLEENVIRONMENT

PLUME

SAMPLING ! AREAA

(Andricevic and Cvetkovic, 1998)

Figure 1 Schemaric o f plume migrating through heterogeneous media


CHAPTERS

METHODOLOGY

The traditional approach to modeling solute transport is based on partial

differential equations which are solved either numerically or analytically to predict

resident concentrations, for a given set o f aquifer parameters and boundary conditions. If

data were available to identify distributions o f aquifer parameters with certainty, the set

o f equations could be solved deterministically: either with numerical or analytical

methods. However in field situations, this level o f data is seldom available, and aquifer

parameters have to be described statistically, yielding to the stochastic formulation of

contaminant transport.

The stochastic approach recognizes that due to natural geologic variability and

scarcity o f data, values o f the hydraulic conductivity and porosity, for example, are not

known throughout the domain, making accurate pointwise predictions o f concentration

d ifficu lt i f not impossible. Assuming aquifer parameters can be described realistically by

a random space function, the best predictors o f solute concentration are the ensemble

mean concentration (C) and the concentration variance, <j^. This thesis uses a solute flux

approach (Dagan et al., 1992) to predict solute transport breakthrough curves at a

compliance boundary for the purpose o f exposure assessment studies (Andricevic and

Cvetkovic, 1996). Solute flux can be defined as the mass o f solute crossing a control


6

plane normal to the mean flow direction per unit time and area. This approach

incorporates significant groundwater flow processes such as advection, macrodispersion,

and mass transfer o f chemical or physical origin.

By assuming hydraulic conductivity is lognormally distributed (Y = In K), its

distribution can be described by the use o f three statistical parameters: the mean value,

Py, the correlation length. Ay, and the standard deviation, oy. The correlation length. Ay,

describes the length over which measurements o f Y are correlated and can be described

mathematically by the following equation:

X y = j p y i ^ d x (1)

where py is the autocorrelation function and x is the separation distance between two

measurements. jUy, and <Ty simply describe the mean and standard deviation o f Y,

respectively. Field observations by Freeze (1975) and Hoeksema and Kitanidis ( 1985)

tend to confirm that Y is a normally distributed population, which is also assumed to be

stationary for simplicity. Stationarity implies that values o f /ty , Ay and oy do not vary in

space throughout the region being studied.

For the objective o f risk assessment it is important to differentiate between the

absolute and relative dispersion (Andricevic and Cvetkovic, 1998). Absolute dispersion

quantifies plume spreading due to velocity fluctuations caused by heterogeneity o f scales

that are both larger and smaller than the plume size. Figure 2a represents a temporal

concept o f absolute dispersion. This figure depicts several o f many equiprobable

contaminant breakthrough curve realizations at a control plane that may result from a

slug release. Due to geologic heterogeneity, it is impossible to predict which to these


7

potential realizations w ill actually occur in the subsurface. Therefore, averaging or

taking the ensemble o f all possible realizations may be o f some value. However, the

resulting ensemble realization w ill have a greater spatial extent than any o f the individual

realizations; therefore effectively increasing dispersion. To avoid artific ia lly increased

dispersion, a relative dispersion concept, which models the motion o f particle pairs, is

suggested for exposure assessment studies, due to potential exposure like ly being caused

by a single realization o f solute transport and its fluctuation. Figure 2b represents a

temporal concept o f relative dispersion. The relative dispersion approach essentially

superimposes each breakthrough realization using its center o f mass. The ensemble of

these superimposed realizations is then calculated which represents the mean

concentration distribution and represents the most likely plume realization that may occur

in the subsurface.

Advantages o f the relative dispersion approach include: ( 1 ) the spatial correlation

structure o f a contaminant plume is predicted which is important in groundwater

monitoring design decisions and (2) a rtific ia lly increased plume dispersion due to

meandering is removed by having the plume centroids coincide. The relative dispersion

approach was used in this thesis.

Lagrangian Transport Formulation

Advective transport is quantified using the Lagrangian framework (Taylor, 1921).

The in itia l solute body or plume is regarded as a collection of indivisible particles that

move in space due to advection by groundwater flow. In relation to groundwater flow the

particles are large compared to the pore scale, but much smaller than other scales

characterizing the plume or the formation (Dagan et al., 1992). A solute particle is


Mus flux snssmUs m un (dueribu abwlut* disparsion)

X

0O Msan trm i Ums in ■

TIME

Mau flux snaamWa mun afiar auparimpming maan traval tlmu (du eribu ralativa dlaparaion)

X

gSs

0

# EiMSinaisnMan iravaillnw0 TIME


Figures 2a (top) and 2b (bottom) Potential breakthrough curves at a control plane


viewed as an indivisible body o f mass that moves along a trajectory toward a control

plane as described in figures 3a and 3b in absolute coordinates and coordinates relative to

the center o f mass, respectively.

Assuming an injected solute mass o f areal density po(a) [M /L ] is advected by a

random groundwater velocity field (V) from a line source at location a, its transport can

be described by the travel time, t = T(x;a), (x is defined in section 3.2) at which a parcel

crosses the control plane. This travel time is obtained by follow ing the advection o f

indivisible solute particles o f conservative mass as they cross the control plane with

transverse coordinates o f intersection y = n(x;a) (Figure 3) (t] is defined in

section 3.2). It is assumed that the solute particle moves in the direction o f the mean

flow and r is positive and finite.

cmlarof RMti Salju

Figures 3a and 3b Particle displacement sketch



10

Kinematical Relationships

For the purposes o f this thesis, solute transport is described using the Lagrangian

time-space process z(x;g) is a time processes which quantifies the travel time o f a

particle transported from the origin, a, to a control plane at .r. q (x:g) is a space process

that quantifies transverse displacement. For a single particle the travel time (x) is

obtained explicitly for the two dimensional case as (Shapiro and Cvetkovic, 1988);

d4

where V/ is the velocity in the mean flow direction.

The component o f the transverse vector for a single particle for the two dimensional case

can be evaluated from the velocity fie ld as:

f ' V, (^, TJ)

V i(4.n)

where V: is the velocity perpendicular to the mean flow direction.

An areally averaged travel time characterizes the plume o f advecting particles in a single

realization originating from a planar source o f area Aq is defined as:

T (x) s X{x\q)da (4)A)

and represents the travel time o f the plume centroid. Sim ilarly, displacement in the

transverse direction is defined as:

Y (%) S ^ f r}(x;q)dq (5)— A)

Therefore, T(x) and Y(x) are the center o f mass coordinates o f a plume at the control

plane.


CHAPTER 4

SOLUTE MASS FLUX

The general mass balance equations for a solute undergoing mass transfer

reactions considering advection only using the flu id mass balance equation are;

(6)

d N

where C is the mobile and N is the immobile solute concentration, and and I/™ denote

the corresponding sink/source terms. Cvetkovic and Dagan ( 1994) presented a solution

to these equations in terms o f a time retention function, y ( t,T ), by coupling solute

advection w ith linear mass transfer reactions.

Some type o f sampling method may detect solutes migrating through an aquifer.

The solute is assumed to be completely mixed within the sample area, denoted by A.

Integrating the solute flux for a single particle, over the injection area, A(y), centered at

Y(y,z) w ithin the control plane, yields the solute mass flux component orthogonal to the

control plane at x as (Dagan et al., 1992):

q{t,y,x,A) = f Po(a)r(f,T) 8 { y ' - f ] )d y 'd a (7)— — —

11


12

where

T{t,z) = (8)

and (j»(t) is the injection rate. The retention time function 7(t,T) is the solution for a solute

pulse o f the advection-reaction system (6) transformed onto an advection flow path

(Cvetkovic and Dagan, 1994). For an instantaneous release o f a nonreactive solute (8)

yields T = 8(t - T). The concentration (C) obtained from (6) and the solute flux for a

single particle, Aq, are related as Aq = C V%(x,q,Q.

Mean and Variance

Taking the expected value o f q (7) yields the mean solute flux (Dagan et al.,

1992):

{‘?(L£)) = “ U o po(fl) rit,r)f;{r,^,x,q)dTd/dq (9)

where f i ' ( x ,y ;x ,q ) is a jo in t probability function which is assumed to follow a log

normal shape for travel time and a Gaussian distribution for transverse displacement.

The assumption o f a Gaussian distribution for transverse displacement is consistent with

numerical simulations in heterogeneous media (e.g. Beilin et al., 1992). The moments

needed to evaluate /," are derived analytically using first order approximation and are

based on the statistics o f a single particle pair. Details o f moment derivation can be

found in Andricevic and Cvetkovic (1997). Omitting explicit dependence o f (q) on Ao

and A, the mean solute flux for a non-reactive solute pulse injection w ith point sample

detection area (i.e. A—>0) is:


13

{'7 (L i)) = f P o (£ )/r { t ,y \x,q)dq (10)—

the variance o f the solute flux is evaluated as

(T;(t,%) (11)

where

{q- (t,x)) = {a)Po(b) F ( t ,yx ,q ,b )dqdb (12)

and

f \ r \ y i t , T ) n t X ) f : {T ,T \y ' , y " \ x ,q ,b )d rd r 'd y 'd y ' ' ( 13)"A * *0 *0 '

where (T, t ' , y ', y"',x,q,b) is the jo in t probability density function for two particle

pairs. The necessary cross-terms are derived in Andricevic and Cvetkovic ( 1997).

For a non-reactive solute pulse w ith sampling over a point (A —>0) the variance o f

solute flux is evaluated as:

( 9 ' ( L £ ) ) = r r Poia)Po(t) fz (^d,y ,y ,a ,b )dqdb ( 14)A) • A> — —

For a non-reactive solute injected over a line source o f extent H, with constant p and a

sampling window o f length B centered at y, the mean solute flux is defined by Andricevic

and Cvetkovic (1997) in two dimensional form as:

{q(t,x,y)) = ^ f ; { f , x , H ) j ^ f i ' { y ' , x , H ) d y ' (15)

the solute flux variance is evaluated as (y^(t,x) = { q ' ) - { q ) '

where


14

{qHt,x ,y)) = ~ i ^ ^ j J ^ j J j : { t , t ; x , H , a ^ ) f : { y \ y " \ x , H , a J d y ' d y " d a ^ d a ' ^ (16)

w ith Oy = I Uy - a'y I is a separation between two particle pairs and M = poH.

Effect of Sorption and Decay

I f solutes undergo equilibrium or non-equilibrium mass transfer reactions between

the fluid phase and surrounding aquifer material, the form o f the time retention function

must be specified. Assuming the sorption-desorption reaction o f the migrating solute is

controlled by first-order kinetics w ith

( p „ , = - a { K j C - N ) - k o C

(17)

i f > , „ = a ( K , C - N ) - k ,N

the time retention function for reversible mass transfer under equilibrium conditions is

(e.g Cvetkovic and Dagan, 1994):

y = 0 [ t - { \ + K , ) x ] (18)

where

ko accounts for irreversible mass transfer

/Td is the partition coefficient

The follow ing equations yield the mean and variance o f solute flux respectively in

two dimensional form assuming the reactive solute is injected over a line source o f extent

H w ith constant po as developed by Andricevic and Cvetkovic (1997)

(< 7 (/,.r,y ))= -^ j^ |j y{t ,T) f ; ( ,X \x)f {{y ' \x )dXdy' (19)


15

the variance o f the solute flux is evaluated as <T U t ,x ,y )= {q ~ ) - {q Y

where

( r u .x .y } ) = j j , J X f y ( r , r ) r ( ! , r ')

f,'{T ,T'\x,H,oc)f,'i y \ y " ,x ,H ,a Jd rd r'd y 'd y "d a ^d a [ (20)

and Oy = I ay - a'y I . A more detailed derivation o f the equations for relative dispersion

can be found in Andricevic and Cvetkovic (1997).

Conversion of Solute Flux to the Flux-Averaged Concentration

The research code (Massflux 2-D), developed by the Desert Research Institute,

used in this thesis predicts a contaminant’s solute flux. In order to relate the solute flux

to water quality standards, the solute flux must be converted to a flux-averaged

concentration. Flux-averaged concentration ( Cf ) can be defined by the following

equation:

\ q v d AC, (21)

j nU -vdA

where:

q is the solute flux representing mass o f solute per unit time and unit area

u i s a unit normal vector

n is the effective porosity,

U is the seepage velocity

and integration is over a planar area, A.


16

The two dimensional code integrates solute flux over a unit thickness o f the aquifer. To

arrive at a flux-averaged concentration (mass/volume), a finite aquifer depth or zone o f

complete m ixing is assumed. Using a finite aquifer depth (or depth o f mixing) a

groundwater flux through the sampling area can be calculated as:

gw flux =U x n x h x B (22)

where:

h is the thickness o f the aquifer (L)

B is the sample detection window (L)

U is the mean groundwater velocity (U T), and

n is the effective porosity


CHAPTER 5

CASE STUDY

From the 1950s to the early 1990s the Nevada Test Site (NTS) was used for the

testing o f nuclear devices. The site occupies 1,350 square miles o f south-central Nevada

and is surrounded on the east, west, and north by the Nellis A ir Force Base Bombing and

Gunnery Range (figure 4). Including the Tonopah Test Range, a 15 to 64 mile buffer

zone exists between the test areas and public lands (NRAMP, 1996). The U.S.

Department o f Energy ( 1994) reports that 815 underground nuclear tests have been

conducted at the NTS. O f these tests, approximately 250, including the 19 largest, were

conducted near (w ithin 100 meters) or below the water table as shown in figure 5

(NRAMP, 1996). This thesis w ill focus on the Benham event, which was detonated 1463

m below ground surface and 822 m below the static water table in borehole U-20c on

December 19, 1968. Benham, one o f the largest underground events, was located on

Pahute Mesa in Area 20 o f the NTS (figure 6) (U.S. DOE, 1994) and had w ith an

announced yield o f 1.15 megatons. This event was selected for investigation due to its

large magnitude, relatively short distance to potential offsite receptors and high potential

for offsite radionuclide migration (Brikowskl and Mahin, 1993). The physical

characteristics pertaining to the Benham event have been summarized in Table 1. The

17


18

Project Shoal Site

Centra) Nevada y Test

Tonopah Test Range _ j

Nclhs Bombtng and Gunnery

Range

Nevada Test Site

Figure 4 State o f Nevada map


19

- y

Nellis Bombing and Gunnery Range Boundary

YuccaMountain

Kikimeten

4-4- ^Frenchman Rat

R o c l^ \ Valley \

I I Nellis Air Fort» Range

0 > 500 kt

# > 200 let

> 20kt

< 20 kt

Figure 5 Location o f detonations near or below the water table


20

N T S B o u n d a ry

Benham

10kmboundary

20kmboundary

V j N e l l i s B o u n d a ry

Oasis Valley boundary (28km) (assumed discharge point)

Figure 6 Location o f Benham event


21

Table 1 Physical Setting of Benham Site

Location Nevada StateCoordinatesDetonated

Announced yield

Surface elevation

Total depth, U-20c

Hole diameter

N903.296 ft,E556,215 ft

December 19, 1968

1.15 megatons

1,914 m

1,463 m

1.8 m

Working point (WP) depth 1,402 m

Rock Unit encountered at WP Zeolitized bedded tu ff

Static water level 1,273 m

Surface distance to major fault 1,750 m (West Boxcar Fault)

Surface topography

Major Stratigraphie Units (below water table)

Flat for a radius o f ~ 1000 m;drops o ff to south-facing c lif f -2,400 msouth o f U-20cTiva Caynon T u ff (Tpc)Bedded Tuffs o f Paintbrush (Tbp) Tonopah Spring T u ff (Tpt)Calico H ills Formation bedded tuffs Tac(b) and lavas Tac(l)______________

(Brikowski and Mahin, 1993)

assumptions be made about critical parameters.

Hydrogeologic Setting

The NTS is located within the southern Great Basin region, an internally drained

part o f the Basin and Range physiographic province (Laczniak et al., 1996). The major

water-bearing units o f this site are (1) the basement confining unit, (2) the carbonate-rock


22

aquifer, (3) the Eieana confining unit, (4) the volcanic aquifers and confining units, and

(5) the valley f ill aquifer (Laczniak et al., 1996). The units o f interest for this

investigation are the volcanic units o f Pahute Mesa that is part o f the Silent Canyon

Caldera, a structural depression resulting from volcanic caldron subsidence

(Blankennagel and Weir, 1973). There are three basic lithologies present in the volcanics

at Pahute Mesa; rhyolitic lavas that tend be more permeable, welded tuffs which are

moderately fractured and permeable, and bedded tuffs which are generally unfractured

and relatively impermeable (Brikowski and Mahin, 1993). According to Blankennagel

and W eir (1973), fractured rhyolites and welded tuffs constitute the important aquifers in

western and central parts o f the caldera, w ith zeolitized ashflall and non-welded ashflow

tuffs acting as aquitards. Although Benham was detonated in a zeolitized bedded tu ff

unit, Brikowski and Mahin (1993) predict the bulk o f the radionuclides produced w ill

move through the lower rhyolite aquifer w ith minimum potential for vertical groundwater

flow through the rubble chimney. Brikowski and W eir (1993) estimate this formation to

be 230 m thick in borehole U-20c. This aquifer is assumed to be continuous and have a

constant thickness o f 200 m for all scenarios. Groundwater flow in lava-flow aquifers is

assumed to be prim arily through interconnected fractures (Laczniak et al., 1996) with the

mean direction o f flow toward the tip o f Oasis Valley (figure 6). DOE ( 1997) has also

suggested this general flow direction.

Source Term

Upon detonation o f a nuclear device, a ring-shaped hydraulic mound is believed

to be created, although the near fie ld environment is poorly understood, it is like ly that

there is a time lag before radionuclide migration would occur form the detonation cavity.


23

Accurate prediction o f chimney in -fill rates requires detailed knowledge o f rock

hydraulic characteristics (Borg et al., 1976). Additionally, the differences in the

estimated effective porosity between the detonation cavity (0 = 0.20) and the surrounding

matrix (0 = 0.005) w ill result in a release o f solutes from the cavity w ith groundwater

flow over some finite time. These phenomena w ill have a significant effect on the

predicted breakthrough curves, particularly for short-lived species. However, due to

scarce near field data and prompt injection o f some radionuclides beyond the detonation

cavity, these phenomena have been neglected and the radionuclide source has been

conservatively modeled as a pulse injection.

Hundreds o f radionuclides are released into the environment upon the detonation o f a

subsurface nuclear device. The generation o f these radionuclides come from four

processes: (1) tritium production in the core o f a thermonuclear device, (2) production

from the fission reactions (fission products), (3) production from the neutron activation o f

surrounding material (e.g. canister and soil) (activation products) and (4) incomplete

fission o f in itia l actinides in the fuel and their transmutation during detonation (NRAMP,

1996). The total mass o f individual radionuclides after an average subsurface nuclear

device is currently an area o f research at the Harry Reid Center (Hechanova, 1997). In

summary, the process used by Hechanova (1997) consists o f :

1. Estimating the in itia l masses o f radionuclides having half-lives greater than one year

present from a nuclear detonation (using publicly available data).

2. Calculating the concentration o f each radionuclide assuming that the mass o f each

radionuclide was completely dissolved in one cavity volume o f water (assuming

cavity has a porosity o f 1.0)


24

3. Comparing this calculated concentration to threshold values (see table 2).

The radionuclides with calculated concentrations greater than threshold values are

listed in Table 2. This table gives the values used to calculate the radiologic source term

which represents the total mass o f the species present. The hydrologie source term is the

mass o f radionuclide that is available for hydrologie transport and may be considerably

less than the total radiologic source term. Smith and Thompson (1995) discuss the

uncertainties in defining hydrologie source terms and cite the follow ing as important

factors:

1. lack o f representative sampling o f cavity waters

2. incomplete knowledge o f radiologic source terms

3. strongly sorbing species

4. inaccuracy in quantifying the transfer process from a radiologic to a

hydrologie source

The upper 95% level o f yield values from Table 2 were used to calculate radiologic and

subsequent hydrologie source terms for the Benham event

Threshold Values

Threshold values for each radionuclide were determined using EPA’s National

Primary Drinking Water Regulations (20 CFR 141). These regulations state that the

average annual concentration o f radioactivity shall not produce an annual dose equivalent

to the total body or any internal organ greater than 4 millirem/year. The following

radionuclide threshold values have been set by the EPA: (20,(X)0 pCi/L), ^°Sr (8


25

pCi/L) and alpha emitters (15 pCi/L). The remaining threshold values were determined

using tissue- or organ-specific dose conversion factors based upon an International

Commission Report (ICRP-30, 1978) and EPA’s Federal Guidance Report No. 11

(Ekerman et al., 1988) and assuming 2L/day ingestion o f contaminated drinking water.

Threshold values for individual radionuclides are listed in Table 2.

Initial Distribution of Radionuclides

Radionuclide distribution and amount available for transport varies depending on

the mechanism by which the radionuclides were created and the chemical characteristics

o f the element. Smith (1993) estimates fission products to be 65-80% incorporated in

melt debris, while activation products were identified as being “ largely” incorporated in

melt debris (NRAMP, 1996). Research shows that less that 1% o f tritium is incorporated

into melt glass. Chemically refractory contaminants however, are approximately 95%

incorporated into melt debris (NRAMP, 1996). A ll radionuclides are assumed to be

evenly distributed throughout a 1 cavity radius w ith the exception o f Tritium and fission

products which are believed to have gaseous precursors and a greater distribution o f 3

cavity radii due to prompt injection (Smith, 1993). Near cavity assumptions are

summarized in Table 3.


26

Table 2 Radiologic Source Term

Radionuclide Half-life(years)

Mean Yield (Ci/kt)

Range of 95% Confidence

(Ci/kt)

Source of Production

ThresholdValue

(pCiÆ)*“ Ru 1.02 1900 400, 3300 Fission 20.9""Sr 28.78 90 54, 150 Fission 8'3?Cs 30.07 140 120, 170 Fission 98.7- ’̂Pu 24100 1.5 0,3.0 Actinide 15238pu 87.7 0.67 0, 1.3 Actinide 15‘^Sb 2.75 290 19,530 Fission 235-> u 6560 0.42 0, 0.84 Actinide 15'•*>m 2.62 580 470, 620 Fission 467-■“ Am 433 0.36 0, 0.72 Actinide 15-■*'Pu 14.4 14 0, 28 Actinide 15

12.32 10000 10, 26000 Tritium 20,000-■“ Cm 18.1 0.33 0,0.66 Actinide 15'^Cs 2.06 2.5 0.002, 16 Fission 63.8

16.1 210 150, 280 Fission 1010'^ tu 8.59 7.8 0.39, 150 Activation 82.3'%Eu 13.5 7.4 0.37, 150 Activation 148'"Eu 4.75 24 6.3, 37 Fission 429

Il3mcd 14.1 0.50 0.02,0.082 Fission 2.6370 0.023 0, 0.046 Actinide 15

- y p 2.14e6 0.0014 0, 0.0028 Actinide 15'"Sm 90 5.9 3.1,6.8 Fission 1470233u I.59e5 0.0098 0, 0.020 Actinide 15-■•*u 2.46e5 0.0081 0,0.016 Actinide 15-■*-Pu 3.75e5 0.00024 0,0.00048 Actinide 15■“ K 1.27e9 0.022 0.0011,0.44 Activation 270

'""Eu 36 0.052 0.0026, 1.0 Activation 237-•‘ 'Am 7370 0.00011 0,0.00022 Actinide 15®'Ni 100 0.33 0.17, 6.6 Activation 1613

'-""Sn 55 0.059 0.001,0.42 Fission 331■‘ 'Ca 1.03e5 0.1 0.005, 2.0 Activation 369'‘c 5720 0.043 0.0021,0.86 Activation 2630

■“ Nb 20000 0.011 0.00055,0.22 Activation 118'-"Sn 2.5e5 0.0036 0.0003,0.0062 Fission 34.2

2.13e5 0.02 0.017,0.022 Fission 437'29i 1.57e7 .00011 4.1e-5,0.00017 Fission 0.6'"Cl 3.01e5 0.013 0.00065,0.26 Activation 133023«u 4.47e9 0.00028 0,0.00056 Actinide 15:'"U 2.34e7 0.00025 0,0.00050 Actinide 15

■ bCmHo 1200 0.0029 0.00015, 0.058 Activation 132235u 7.04e8 0.00014 0,0.00028 Actinide 15

'■“ Tm 1.92 0.010 0,0.015 Fission 1130“'Zr 1.5e6 0.0023 0.0016,0.0032 Fission 162


27

Table 3 Assumptions for Near Cavity Distribution of Radionuclides

Category Percent Incorporated Into M elt Debris

Distribution Radius Factor

Tritium 0 3 cavity radii

Fission Products 70 3 cavity radii

Activation Products 70 1 cavity radius

Actinides 95 1 cavity radius

(NRAMP. 1996)

Solubility of Radionuclides

The solubility o f radionuclides varies according to the chemistry o f the element,

water temperature, and groundwater chemistry (Kerrisk, 1984). Solubility constraints

may be important considerations due to the potential to restrict the mass available for

transport. Table 4 gives solubility estimates o f some important radionuclides using water

from NTS well J-13 to define the water chemistry (values are an average o f six analyses).

W ell J-13 is developed in a volcanic unit o f the southwestern part o f the NTS. Using

solubility estimates (table 4) and an assumed distribution, the individual actinide

hydrologie source terms could not be reduced due to solubility lim its. Solubility values

presented were estimated for each radionuclide alone; the results are, therefore, the upper

lim its o f solubility because many radionuclides can compete to form complexes with the

same components o f groundwater. This effect may be important among a group o f

chemically sim ilar elements such as the actinides (Ogard et al., 1984). Solubility

estimates vary greatly as illustrated by Ogard et al., ( 1984) who suggest the solubility o f


28

Pu may be as low as 1.4 e -17 M. It is understood that variability in water chemistry and

temperature w ill greatly affect solubility values.

Table 4 Solubilities of Selected Radionuclides

Element Solubility (M)

Am 1.08e-8

Pu 1.8e-6

U 2.1e-4

Sr 9.4e-4

Sn l.Oe-9

Cs large

Tc large

Np 3.0e-3

(Kerrisk, 1984)


29

Table 5 Well J 13 Water Chemistry Data

Species Total Amount (mg/L)

Na 45.2

K 5.47

Ca 11.5

Mg 1.73

A1 0.026

Si02 64.2

Sr 0.04

Ba 0.0023

Mn 0.0011

Fe 0.044

V 0.0425

Cl- 6.4

SO4- 18.1

F 2.1

NO3 10.1

HPO4- 0.10

HCO3 143

( pH = 7; Eh = 700mV, temperature was not stated) (Kerrisk, 1984)


30

Spatial Variability in Hydraulic Conductivity

Hydraulic conductivity (K) varies spatially due to the natural variability o f

geologic formations. The variability in K produces tortuous and unpredictable flowpaths

which have both faster and slower velocities than those calculated using the mean K

value. These faster and slower flowpaths result in plume spreading along the direction o f

flow. The magnitude o f this variability can be quantified by the variance o f InK. A large

variance w ill increase simulated spreading relative to the mean value, while an small

variance w ill restrict the spreading about the mean. This variance is o f particular

importance when considering decaying solutes o f relatively short half-lives. A large

variance may result in arrivals earlier than those o f the mean and therefore higher

activities due to less radioactive decay. Hoeksema and Kitanidis ( 1985) analyzed 16

transmissivity measurements from the fractured rhyolitic lava in Pahute Mesa and

reported a structured variance o f ln(transmissivity) o f 0.276, and a median value o f 0.30

for all consolidated aquifers analyzed. Hoeskema and Kitanidis (1985) do not suggest a

value for variance of InK in consolidated aquifers but cite a median value o f 0.56 for the

variance o f InK in unconsolidated aquifers. They also suggest that consolidated aquifers

generally display smaller structured transmissivity variability than unconsolidated

aquifers, suggesting that the smaller variability is due to the greater uniform ity over large

distances. Due to a lack o f hydraulic conductivity data between the contaminant source

and the control planes, a value o f 1.2 was assumed for the variance o f In K.


31

Source Size

The follow ing equation was used to calculate the cavity radius (Associated Rocky

Mountain Universities, 1965). The same equation was used by the DOE/NV in the

Underground Testing Agency (UGTA) modeling program (DOE/NV, 1996) and by

NRAMP (1996).

C xW '" (p xh) '= ______ . , 1/4 ( - 3 )

where:Rc = cavity radius (m)C = constantW = device energy yield (kT) p = average overburden density (g/cm ') h = depth o f overburden (m)

A value o f 70 was used for C, the mean o f the Associated Rocky Mountain Universities

studies (1965). This value was also used by DOE/NV in the UGTA. Based on a grain

density o f 2.65 g/cm' and 17 percent porosity, the average overburden density was

calculated to be 2.2 g/cm' (NRAMP, 1996). This value is expected to vary due to

geologic heterogeneity but the cavity radius equation is not highly sensitive to this

parameter. Actual cavity radius w ill vary depending upon yield emplacement geometry,

and characteristics o f the host rock such as water content and strength properties (Borg et

al., 1976). Direct measurements show that cavities have significant irregularities and that

radii vary by as much as 15% along different transects (Laczniak et al., 1996). Using

equation (23), a value o f 100 m was calculated for the cavity radius o f the Benham event.

This value was used along w ith the distribution radius o f each contaminant to arrive the


32

source size used in modeling scenarios. A sensitivity analysis o f concentration to source

size was performed and is presented later is this thesis.

Sampling Detection Window

A sampling window o f 100 m is assumed. This value represents a hypothetical

capture zone induced by a well o f small discharge. I f a larger discharge is assumed, or

more than one well were pumping, this value would increase. Variation o f sampling

window has only a minor effect on the mean solute flux. The variance o f solute flux,

however, is more sensitive to the sampling window, with variance o f solute flux

decreasing as sampling length is increased (Andricevic and Cvetkovic, 1998).

Correlation Scale

The correlation scale is the distance at which two measurements tend to become

uncorrelated. Dagan ( 1984, 1986) suggests two fundamental correlation scales: the local

and regional scales. The local scale refers to the spatial variation o f the hydraulic

conductivity in a domain whose size is o f the order o f the aquifer depth in the vertical

direction and o f the same order in the horizontal plane (Dagan, 1988). The local

correlation scale has been found to be on the order o f meters. The regional scale refers to

the entire aquifer whose horizontal extent is much larger than the depth. A t this scale,

flow variables are averaged over depth, and the flow is viewed as two-dimensional in the

horizontal plane. A large correlation value suggests high spatial correlation, effectively

extending the length o f higher conductivity flowpaths (NRAMP, 1996). There is little

data available to determine an accurate correlation scale for the scenarios modeled. For


33

consistency, a value o f 200 m was assumed for the regional correlation scale for all

scenarios.

Aquifer Thickness

The thickness o f the lower rhyolitic aquifer is 230 m thick at the location o f the

U-20c borehole (Brikowski and Mahin, 1993). Blankenagel and Weir (1973) cite

thickness o f 0-518 m for the lower rhyolite lava in the western part o f the Silent Canyon

caldera. Due to the lack o f formation thickness data between the source and control

planes, a constant aquifer thickness o f 200 m was assumed.

Radionuclide Retardation

Sorption unto the host matrix may retard the transport o f some radionuclides. The

magnitude o f sorption can vary depending on groundwater chemistry, particle size,

radionuclide o f interest and abundance and distribution o f sorbing minerals along

transport paths. Much o f the geology o f the NTS is composed o f Miocence and younger

ashfall and ashflow tuffs which contain highly sorptive minerals (Smith, 1993). A

radionuclide’s a ffin ity to sorb can be quantified under equilibrium conditions by its

distribution coefficient (^a):

Kd = (amount o fX within solidVfweieht o f solid)(amount o f X within liquid)/(volume o f liquid)

where X is a specific radionuclide. values can be determined in the laboratory with

magnitudes depending on the chemical composition o f the liquid phase, the composition.


34

particle size and texture o f the solid phase, temperature and related environmental

parameters (Smith, 1993). Isherwood ( 1977) further emphasizes the role o f sorption in

mitigating radionuclide migration by defining the retardation factor (Kf or R) assuming

linear isotherms:

Kf = I + rKd

where = distribution coefficient and

r = ( l -d)p d

where d = effective porosity and p = bulk density. The retardation factor is the ratio o f

groundwater velocity to that o f solute velocity. Retardation may effectively decrease

radionuclide concentrations o f relatively short-lived radionuclides through radioactive

decay and significantly slow the migration o f long-lived species.

Isherwood (1977) evaluated the retardation capacity for 23 radionuclides in

western soil (see Table 6) and suggests minimum and maximum values for retardation.

In Table 7, Kerrisk (1985) gives median retardation values for a variety o f Yucca

Mountain Tuffs and notes that these values can vary over a order o f magnitude depending

on mineralogy. Due to not knowing the flow path mineralogy or pH variations between

the source and control plane, the low range o f retardation literature values were selected

for this thesis.

Increasing evidence exists that indicates field scale sorption rates are lower than

generally assumed and that kinetic effects may be significant (Ptacek and Gillham, 1992),

particularly fo r transport in fractured formations (Knapp, 1989). Physical as well as


35

chemical processes can result in nonequilibrium retardation. M atrix diffusion can retard

radionuclide transport where water flows primarily though fractures, including low or

non-sorbing species. Although nonequilibrium sorption and matrix diffusion are

probable, quantifying data is scarce and as such are not addressed in this thesis.

Table 6 Retardation Factors for Western Soil

Radionuclide Minimum Preferred MaximumIodineTechnetium

1 1 1

Other Fission Products 1 100 1,000

Actinides 100 10,000 100,000

(Isherwood, 1977)

Relevant sorption data for many o f the species in the source term could not be located

and although they may sorb strongly, were modeled as chemically conservative species.

Both tables 6 and 7 neglect the potential for colloidal transport o f radionuclides, which

conversely would increase the rate o f migration for many radionuclides. Retardation

factors assumed for this investigation (Table 8) are only intended to be an order o f

magnitude estimate, species not listed were assumed to be transported conservatively.


36

Table 7 Median Retardation Factors from a Variety of Yucca Mountain Tuffs

Element Retardation Factor

Np 50

U 25

Pu 500

Am 1,000

Cm 100

Th 500

Ra 1,000

Cs 1,000

Sr 700

Tc 1

C 1

I 1

Sn 1,000

Ni 20

Zr 100

Sm 1,000

(Kerrisk, 1‘

Table 8 Assumed Retardation Factors

Radionuclide Retardation Factor

^'Zr, ^'N i, "^Nb 10

^S r, "'C s , "'S m , '- '" ’Sn, '"^Sn, 100

^°K, ■"Ca, '“ "’Ho 1000

(Isherwood, 1977; Kerrisk, 1985; Sheppard and Thibault, 1990)


37

Colloid Facilitated Transport

Laboratory and field evidence exists indicating that colloidal material is mobile in

groundwater (McDowell-Boyer et al., 1986) and that it is possible, especially for low

solubility and highly sorbing radionuclides to be transported by these mobile colloids

(Buddemeir and Hunt, 1988). Three major effects o f radionuclide-colloid interaction on

the transport and breakthrough o f radionuclides as summarized by Grindrod (1993) are:

( 1 ) colloid-bound radionuclides may proceed at a faster flow rate (on average) than

conservative dissolved solute species.

(2) colloids may become immobile, thereby entrapping currently sorbed radionuclides.

The retardation o f radionuclides due to colloid-rock sorption is dependent upon both

the colloid-radionuclide sorption behavior as well as the colloid-rock

sorption/desorption rates.

(3) dispersion o f colloid-bound solutes may be distinct from that o f dissolved solutes.

Studies by Buddemeir and Hunt (1988) at the NTS Chesire event suggest that colloids

in the range o f 0.006 to 0.003 pm sorb radioactive Mn, Co, Cs, Ce and Eu species.

Buddemeir and Hunt ( 1988) and Smith ( 1993) suggest that most NTS are transported less

efficiently than soluble neuUal or anionic species. IT Corporation ( 1997) suggests

colloids with sorbed radionuclides may have IQ values ranging from 0-10 mg/ml (table

9) in NTS alluvium.


38

Table 9 Colloidal Transport Parameters for Frenchman Flat

Radionuclide H alf-life (years) Kd Range (g/ml)

"'C s 30.3 Oto 10

23«J/240pjj 2.4e4 Oto 10

2.3«u 4.5e9 Oto 2

(IT, 1997)

Due to the potential for colloids to be transported as fast or faster than average

groundwater flow, actinides which are suspected to be transported colloidally, have been

modeled as chemically conservative species. Recent analysis o f groundwaters at the NTS

has detected Pu 1.3 km from the Benham event (Thompson et al., 1997). This analysis

confirmed using classified isotopic ratios that the Pu was produced by the

Benham event. Thompson et al. (1997) also suggests that the Pu may have been

transported colloidally. “ Calibrabration” o f model parameters is not possible from this

single data point.

Selection of Control Planes

Uncertainty in future NTS landuses provided the impetus to model transport to

multiple control planes (figure 6). The 28 km boundary represents the distance to the tip

o f Oasis Valley which is the assumed discharge area.

Mean Groundwater Velocity

Blankenagel and Weir (1973) used transmissivity values and an estimated

regional hydraulic gradient to estimate a discharge beneath Pahute Mesa. Using an


39

assumed cross sectional area and an effective porosity range o f 0.20 to 0.005 calculated

groundwater velocities ranged from 2 to 76 m/yr, respectively. To avoid underestimating

solute migration, a value o f 70 m/yr was selected for this thesis. Other model parameters

used are summarized in table 10.

Procedure

Using the assumed hydrologie source term, and the parameters stated in table 10,

all species were modeled to the 10 km control plane. Species with predicted

concentrations above threshold values using the upper 95% level o f peak concentration

were modeled to the next boundary, while species below the threshold values were no

longer considered for the remainder o f the investigation. This procedure was repeated

using distances o f 20 and 28 km for control planes.

Conversion of Flux Averaged Concentration to C. E. D. E.

The predicted radionuclide concentrations were converted to committed effective

dose equivalent (C.E.D.E.) using the species specific threshold concentration as shown in

the example below for ‘‘*C.

assumed concentration = 8.18e5 pCi/Lthreshold value = 2.63e3 pCi/L = 4 mrem/yr (C.E.D.E)

8.18e5C.E.D.E. = — x4 = 1244 mrem/ vr

2.63e3

This conversion assumes a linear exposure to dose relationship and may not be valid for

low doses.


40

Table 10 Parameter Summary

PARAMETER VALUEAquifer Rhyolitic lava (200m thick)

Variance o f In K 1.2

Sample detection window 100 m

Source size 600 m (^H and fission products)200 m (activation products and actinides)

Decay rate radionuclide specific

Retardation factor radionuclide specific

Correlation scale 200 m

Mean groundwater velocity 70 m/yr

Effective porosity 0.005


CHAPTER 6

RESULTS

Tables 11 through 19 present the fission products (tritium, although distinct, was

grouped with the fission products for convenience), activation products and actinides

which have predicted concentrations above their threshold values at 10, 20 and 28 km

boundaries using the upper 95% confidence level o f peak flux-averaged concentration.

The majorities o f radionuclides present from the Benham event do not have relevant

sorption data or have the potential for colloidal transport, these radionuclides were

modeled as chemically conservative species, though many are expected to sorb strongly.

Fission products that have predicted 95% confidence level of peak concentrations above

threshold values at Oasis Valley are '"’ l, and ’ ^Tc (table 17). Activation products

predicted to be above threshold values at Oasis Valley include '^®Eu, ^ K , ’‘‘C, ‘̂*Nb, and

®̂C1, and (table 18). Table 19 presents the 13 actinides predicted to be significant at

Oasis Valley. A ll actinides were assumed to have the potential to be transported

colloidally. Buddemeir and Hunt (1988) recognized the potential for colloidal transport

o f radionuclides at the NTS but suggest that colloids are transported less efficiently than

anionic or neutral species. Additionally, the IT Corporation ( 1997) reported a

distribution coefficient range o f 0-10 mg/L for colloidal transport o f (table 9) in

NTS alluvium. However, due to the potential for nonsorbing colloidal transport, all

41


42

actinides were modeled as chemically conservative species. The assumptions made in

this thesis regarding colloidal transport and solubility lead to predicted concentrations

o f actinides which are believed to be unreasonably high. This is illustrated by the

comparison o f predicted concentrations at Oasis Valley to the maximum observed NTS

actinide groundwater concentrations published (table 20). The values in table 20 are

from a Lawrence Livermoore National Laboratory data base which includes data from

impacted wells near or within test cavity areas (NRAMP, 1996).

It is obvious that although values in table 20 may not be the maximum

concentrations possible, the large disparity between these and predicted values strongly

suggests both physical and chemical refinements regarding actinide transport is needed.

For this reason no further analysis o f actinide results are presented. Predicted

concentrations of the significant fission and activation products at Oasis Valley were

converted to committed effective dose equivalent (see section 5.17). Figure 7 shows the

breakthrough curves o f the upper 95% confidence level of committed effective dose

equivalent from significant species at Oasis Valley. It is important to note that although

the 95% level o f committed effective dose equivalent for and '^^Eu are above

threshold levels, the mean values are not. This figure shows the temporal relationship

o f the breakthrough for each species (retardation factors of 1, 10, and 1000 were used).

Retardation may not result in reducing concentrations o f long-lived species below

threshold levels, but may significantly delay their breakthrough. This is depicted in the

figure which shows that the 4 mrem/yr standard is not predicted to be broken until 3,500

years after detonation for ^̂ *Nb, while the standard is not predicted to be broken for

260,000 years by Figure 8 shows the 95% confidence level o f committed effective

dose equivalent breakthrough for the assumed conservative species. The 4 mrem/yr


43

Standard is predicted to be broken at times ranging from 290 to 350 years post

detonation for and respectively.

Table 11 Fission Products Above Threshold Values at 10 km

Species Half-life(years)

Hydrologie source term

(pCi)

Kf Threshold Cone. (pCi/L)

Peak (C> (pCiÆ)

Peak (C) + 2gc

{pCUL)' h 12.3 2.99el9 1 20,000 8.64e6 3.27e7

IL Im C d 14.1 2.83el4 1 2.63 2.22e2 7.02e2.29] 1.57e7 5.87elO 1 .59 23.3 65.5

99tc 2.13e5 7.59el2 1 437 4.18e3 8.47e3

Table 12 Activation Products Above Threshold Values at 10 km



(pCi)


Peak <C> (pCi/L)

Peak (C) + 2Cc

(pCiÆ)8.59 5.18el6 1 8.23el 1.62e3 6.67e3

■'-Eu 13.5 5.18el6 1 1.48e2 8.8 le4 2.7e5

""Eu 36 3.45el4 1 2.37e2 4.88e4 1.21e540^ 1.27e9 1.52el4 1000 2.7e2 3.38e2 7.47e2

■“ Ca 1.03e5 6.9el4 1000 3.6e2 5.75e2 1.33e3

" c 5720 2.97el4 1 2.63e3 6.49e5 1.43e6

"Nb 2.0e4 7.59el3 10 1.18e2 1.61e4 3.55e4

*C1 3.01e5 8.97el3 1 1.33e3 I.99e5 4.4 le5


Table 13 Actinides Above Threshold Values at 10 km

44

Species Half-life(yeai-s)


(pCi)


Peak (C) (pCi/L)

Peak (C) + 2oc

(pCi/L)

24000 1.73el4 1 15 3.83e5 8.47e5

%*Pu 87.7 7.48el3 1 15 5.3le4 1.23e5

- ^ U 6560 4.83el3 1 15 1.06e5 2.34e524lpu 14.4 1.61el5 1 4.26 4.13e3 1.27e4

-■“ Am 433 4.14el3 1 15 7.36e4 1.63e5

-^Cm 18.1 3.8el3 1 15 3.93e2 1.07e3232^ 70 2.65el2 1 15 1.43e3 3.3le3

:"N p 2.14e6 1.61ell 1 15 3.58e2 7.91e2

-” u 1.59e5 1.15el2 1 15 2.55e3 5.65e3234u 2 46e5 9.2ell 1 15 2.04e3 4.52e3

-•*-Pu 3.75e5 2.76elO 1 15 6.13el 1.36e2

-•*'Am 7.37e3 1.27elO 1 15 2.79el 6.16el238u 4.47e9 3.22elO 1 15 7.15el 1.58e2236u 2.37e7 2.88elO 1 15 6.40el 1.42e2235u 7.04e8 1.61el0 1 15 3.58el 7.91el




(pCi)


Peak (C) (pCiÆ)

Peak <C) + 2ctc

(pCiÆ).291 1.57e7 5.87el0 1 .59 2.23el 4.54el

99tc 2.13e5 7.59el2 1 437 2.88e3 5.86e3


45




(pCi)


Peak <C> (pCi/L)

Peak (C) + 2oc

(pCi/L)

"-Eu 13.5 5.18el6 1 1.48e2 4.94el 1.72e2

""Eu 36 3.45el4 1 2.37e2 2.3 le3 5.74e3

40^ 1.27e9 1.52el4 1 2.7e2 2.35e2 5.22e2

'^C 5720 2.97el4 1 2.63e3 4.44e5 9.84e5

^N b 2.0e4 7.59el3 10 1.18e2 1.07e4 2.36e4

‘̂’Cl 3.0 le5 8.97el3 1 1.33e3 1.41e5 3 07e5




(pCi)


Peak (C) (pCiÆ)

Peak <C) + 2gc

(pCi/L)2.39pu 24000 1.73el4 1 15 2.66e5 5.88e5238pu 87.7 7.48el3 1 15 1.2 le4 2.8 le4

:4"Pu 6560 4.83el3 1 15 7.26e4 1.61e524.pu 14.4 1.61el5 1 4.26 3.54 1.23el

:"A m 433 4.14el3 1 15 4.09e4 9.05e4232^ 70 2.65el2 1 15 2.47e2 5.73e2

^^Np 2.14e6 1.61ell 1 15 2.49e2 5.52e2233u 1.59e5 1.15el2 1 15 1.81e3 3.93e3234u 2.46e5 9.2ell 1 15 1.45e3 3.15e3242pu 3.75e5 2.76elO 1 15 4.27el 9.45el

:'"Am 7.37e3 1.27el0 1 15 1.91el 4.24el238^ 4.47e9 3.22elO 1 15 4.98el l.le2236u 2.37e7 2.88eI0 1 15 4.46el 9.87el235u 7.04e8 1.61el0 1 15 2.49el 5.52el


46




(pCi)


Peak (C) (pCi/L)

Peak (C) -1- 20c

(pCi/L).291 1.57e7 5.87elO 1 .59 l.87el 3.8el

^ C 2.13e5 7.59el2 1 437 2.42e3 4.9 le3




(pCi)


Peak (C) (pCi/L)

Peak (C) + 2oc

(pCi/L)

""Eu 36 3.45el4 1 2.37e2 2.05e2 5.59e2

'"K 1.27e9 1.52el4 1 2.7e2 2.02e2 4.39e2

'“C 5720 2.97el4 1 2.63e3 3.75e5 8.18e5

^Nb 2.0e4 7.59el3 10 1.18e2 8.62e3 1.91e4

3.01e5 8.97el3 1 1.33e3 1.17e5 2.59e5


47




(pCi)


Peak (C) (pCi/L)

Peak (C) + 2 g ç

(pCi/L)239pu 24(XX) 1.73el4 1 15 2.27e5 4.94e5238pu 87.7 7.48el3 1 15 4.18e3 9.69e3240pu 6560 4.83el3 1 15 6.03e4 1.34e5

:"A m 433 4.14el3 1 15 2.87e4 6.36e4232u 70 2.65el2 1 15 6.8el 1.58e2

:"N p 2.14e6 1.61ell 1 15 2.14e2 4.65e2233^ 1.59e5 1.15el2 1 15 1.52e3 3.32e32.34U 2.46e5 9.2ell 1 15 1.2e3 2.65e3242pu 3.75e5 2.76el0 1 15 3.66el 7.97el

-^'Am 7.37e3 1.27elO 1 15 1.59el 3.53el238u 4.47e9 3.22elO 1 15 4.27el 9.3el236u 2.37e7 2.88el0 1 15 3.82el 8.32el2.35U 7.04e8 1.61el0 1 15 2. le i 4.65el

Table 20 Maximum Observed Actinide Concentrations

Actinide Concentration (pCi/L) Citation

4J8pu 2.4e-8 LLNL (Fraser 1982)“

239pu 9.3e-0 Hoffman et al. (1977) a

:-Wpu 9.3e-4 LLNL (1995)

-^'Pu 7eO LLNL (1995)

234u 4.9e0 Erikson and Jacobson ( 1990) “

235u

238u

8.6e-l

2.7el

Buddemeir and Isherwood (1985)“Buddemeir and Isherwood (1985)“

cited in Daniels et al. (1993)


48

10,000 -

1,000 -

I111

SÜ

100 -

Breakthrough of 95% Level of Committed Effective Dose Equivalent at Oasis Valley

-1-129

-Tc-99Eu-150

K-40-C-14-Nb-94

-CI-36»4mrem/yr

1,000 10,000 100,000 1,000,000

Tim* (yaars)

Figure 7 Breakthrough o f 95% level of C,E.D.E, at Oasis Valley

Breakthrough of 95% Level of Committed Effective Dose Equivalent at Oasis Valley (conservative species)

10,000 -

1,000

Î 100 -

IsI II

10 •

( j

1 •

0 i250

Tc-99

Eu-150

4 mrem/yr

350 450

Tim* (y*ar9)

Figure 8 Breakthrough of 95% level o f C.E.D.E. (conservative)


49

Although results thus far have concentrated on determining i f and when each

species breaks the 4 mrem/yr threshold, drinking water regulations actually lim it the

combined concentration of radioactivity from producing more than an annual dose

equivalent to the total body or any internal organ o f 4 mrem/yr. Figure 9 shows the

mean and 95% confidence level o f the effective dose breakthrough for the summation

(assumes coincident arrival) o f the five significant assumed conservative species

**’®Eu, and ^^Cl). Using the summation of the 95% confidence level o f

committed effective dose equivalent breakthrough, the 4 mrem/yr threshold is not

predicted to be broken until 280 years after detonation.

Contaminant plumes generally undergo hydrodynamic dispersion as they are

transported by groundwater. This dispersion effectively decreases the resident

groundwater concentrations o f solutes due to increased plume size both laterally and

longitudinally. The method used in this thesis models macrodispersion.

Macrodispersion results from velocity fluctuations due to the spatial variability of

hydraulic conductivity only, neglecting both pore scale dispersion and diffusion. Pore

scale dispersion may significantly increase lateral dispersion at long travel times.

Figure 10 shows the temporal and lateral extent o f the 95% level predicted plume for

'“’ l. This figure illustrates the predicted peak lateral extent o f the 5 mrem/yr plume to

be over 1200m.

Sensitivity Analysis

Due to the lack o f hydrogeologic data along the flowpath from the Benham event to

the assumed discharge location, large discrepancies may exist between model

parameters and actual site conditions. To understand the relative importance of the


50

Summation of Committed Effective Dose Equivalent at Oasis Valley (mrem/yr of conservative species)

1.0E+04 -

1.0E+03 -

f5 1.0E+02 -

uid 1.0E+01 - 4 _O ■

1.0E+00 * '

•MeanC.E.D.E.

- Mean + 2 I Sigma I C.E.D.E. I

•4 mrem/yr

1.0E-01250 300 350 400 450 500 550

Tim* (year*)

Figure 9 Summation o f C.E.D.E. (conservative species only)

800

600 -

400 -

2 0 0 -

mCL CM

-200 -

= -400 -

-600 -

-800250 300 400 450 500350 550

Time (yecrs)

Figure 10 '“’l 95% level C.E.D.E at Oasis Valley (mrem/yr)


51

uncertainty in various modeling parameters, sensitivity analyses were conducted for the

source size, mean groundwater velocity, variance o f hydraulic conductivity and

retardation factor. Parameter sensitivity may vary depending on radioactive half-life o f

solute modeled. The sensitivity analyses were conducted by holding all parameters

constant, except the one being examined. Possible correlation among parameters has

been ignored. Although uncertainty in the source term exists, the variation scales

linearly and, therefore, results are not presented here.

Figures 11 and 12 show the results o f the sensitivity analysis for source size.

Fission products are assumed to be evenly distributed over 3 cavity radii. These figures

show the variation of breakthrough and peak concentration with a source size

distribution ranging from 1 to 4 cavity radii for and '"’ l. Results indicate larger

source sizes have an earlier breakthrough and a lower peak value for long-lived species.

Peak values range over a factor of 6 for as the source size distribution ranges from 1

to 4 cavity radii. Little variation in peak concentrations for were predicted, although

the time to peak concentration was up to 50 years earlier assuming the larger source

size. peak predictions are below threshold concentrations for all source sizes tested

while '*^1 peak predictions are above threshold values for all cases.

Figures 13 and 14 show the results o f the sensitivity analysis for mean

groundwater velocity for ^^Eu (ti/2 = 36 yrs) and (t,/2 = 1.57e7 yrs). Data from

Pahute Mesa analyzed by Blannkenagel and Weir (1973) suggest an estimate of velocity

to be 70 m/yr based on discharge measurements and assumed effective porosity of

0.(K)5. Assuming porosities o f 10 and 2%, low and mid range velocities o f 4 and 20

m/yr were used as sensitivity values. From figure 13 it is clear that peak concentrations


Sensitivity of "̂ 1 Concentrations at Oasis Valley to Source Size Distribution

g 1000

%eO

!Ig

■ 1 radii- 2 radii

3 radii

4 radii

100

10

1

0.1

0.01200 300 400 600500

Tim# (year#)

Figure 11 Sensitivity o f '"^1 concentrations to source size

Sensitivity of Tritium Concentrations at Oasis Valley to Source Size Distribution

100 T

I :I

1 ^Ii

0.1200 250 300 350 400 450

Tim# (years)

-1 radii

-2 radii

3 radii

4 radii

500

Figure 12 Sensitivity o f concentrations to source size


53

of species with relatively short half-lives are highly sensitive to groundwater velocity.

Figure 14 shows that peak concentrations of longer-lived species such as are less

sensitive (varies by over a factor o f 20) but indicate that time to peak concentration

varies greatly.

The results o f the sensitivity analysis for variance o f hydraulic conductivity

((TinK) are shown in Figures 15 and 16. Little site data was available for determining an

accurate value for this parameter. Values of <TinK = 0.60, 1.2 and 1.8 were used as

sensitivity values. It should be noted that the solute flux method is based on first order

analysis which is most accurate when CTinK < 1. These figures illustrate the sensitivity

o f both breakthrough time and peak concentrations for short and long-lived species to

CTinK- A variance o f 1.8 produces earlier breakthrough and a higher peak concentration

in ^H, while the same variance produces an earlier breakthrough but lower peak

concentration for the long-lived '"’ l.

Figures 17 and 18 show the results o f a sensitivity analysis for retardation factor.

Figure 17 shows that ’ °Sr is highly sensitive to retardation factor and the peak

concentration is reduced below its threshold value assuming a retardation factor of 2.

Figure 18 shows the sensitivity o f * ’̂ Pu to retardation. Although all actinides were

modeled as conservative species due to the potential for colloidal transport, it is

believed that the colloids may not be transported as efficiently as conservative species.

Additionally, colloids may not be present in sufficient quantities to contribute

significantly to the transport of actinides. This figure illustrates the sensitivity o f *^^Pu

to ‘effective’ retardation factor. I f colloidal transport is ignored, the retardation factor

may be as high as 100,000 (Isherwood, 1977).


54

Sensitivity of '̂ °Eu Concentrations at Oasis Valley to Mean Groundwater Velocity

A1.E+03 - 1.E+02 - 1.E+01 -

% 1.E+00 - ■5 1 .E-01 ** I E.0 2 . , - * - u = 70m /yr

8 1 .E.0 3 - - * - u = 20m /yr

8Ô 1 .E -0 5 - S 1.E-06 r

J 1 E - 0 7 - S 1.E-08 -

1.E-09 —

1.E-04 - u = 4 m/yr

100 1.000 10.000

Tima (years)

Figure 13 Sensitivity o f '̂ ®Eu concentrations to groundwater velocity (results for u=4 m/yr were below computation limits)

Sensitivity of '̂ ’l Concentration at Oasis Valley to Mean Groundwater Velocity

_ 100 T

I :10

-u = 4 m/yr -u = 20 m/yr

u = 7 0 m/yr

0.1 . . . . .

100 1,000 10.000

Tima (yaars)

Figure 14 Sensitivity o f '*^1 concentrations to groundwater velocity


55

Sensitivity of Tritium Concentrations at Oasis Valley to Variance in Hydraulic Conductivity

100 T

%

!%

Ii

0.1 -------- --------- --------- --------- --------- ---------f —

150 200 250 300 350 400 450 500

■Sigma Y = 0.6 -sigma Y = 1.2 sigma Y = 1.8

Tim* (years)

Figure 15 Sensitivity o f H concentrations to Oy

Sensitivity of '**1 Concentrations at Oasis Valley to Variance in Hydraulic Conductivity

_ 60 -

% 50 -

ou

40 *

30

Ô 20

I10

150

■Sigma Y = 0.6 ■Sigma Y = 1.2

Sigma Y = 1.8

350 450 550Time (years)

650

Figure 16 Sensitivity o f concentrations to Oy


56

Sensitivity of Concentrations to Retardation Factor at Oasis Valley

' i

1.E+06 -

1.E+05 -

1.E+04 *

1.E+03 -

1.E+02 -

1.E+01 ^

1.E+00 -

1.E-01 ■

1.E-02 -

I E 03 -

1.E-04 —10O

■R = 1 -R = 10 R = 100 R=1000

1,000 10,000

Time (years)100,000 1,000,000

Figure 17 Sensitivity o f concentrations to retardation factor

Sensitivity of *°Sr Concentrations at Oasis Valley to Retardation Factor

^ 1.E+04 -

^ 1.E+02 -

f 1.E+00 -

I 1.E-02*

g 1.E-04 go 1.E-06 O% 1.E-08

Ig

1.E-10

1.E-12

1.E-14

■R = 1

-R = 2 R = 5

500 1000 1500

Time (years)

2000 2500

Figure 18 Sensitivity o f ̂ S r concentrations to retardation factor


57

Uncertainty in Mean Plume Position

The case study used in this thesis illustrates an example o f “ non-ergoditic”

transport i.e. plume scale is small or comparable to the heterogeneity scale such that

plume meandering may be significant. As mentioned earlier in this paper, the relative

dispersion method does not account for plume meandering, however, the code used

(Massflux-2D) quantifies the uncertainty in mean plume positions. The difference

between the relative and absolute dispersion variances describes the uncertainty of the

mean plume position in time and transverse displacement. It is important to note that

this method neglects pore scale dispersion and variations in geology such as the

structure and lithology o f the aquifer and only represents the statistical uncertainty in

the mean plume position due to the spatial variability o f hydraulic conductivity. Using

‘■̂1 as a test case, the 95% confidence level o f plume position is within ± 44.8 years

temporally and ± 60 m in transverse displacement o f the mean at Oasis Valley.


CHAPTER 7

DISCUSSION

Twenty o f the initial 42 radionuclides in the hydrologie source term have

predicted peak 95% confidence level concentrations above their respective threshold

values at the Oasis Valley control plane. Tritium, which has historically been a main

concern in contaminant transport studies at the NTS, did not break the 20,000 pCi/L

threshold at Oasis Valley in any o f the scenarios or sensitivity analyses used in thesis. O f

the twenty potentially significant radionuclides, 13 are actinides. The assumptions made

for the transport modeling of the actinides are very conservative. The total hydrologie

actinide source term is assumed to be irreversibly sorbed to colloids which in turn w ill

exhibit no sorption to the host matrix during transport. More likely, only a fraction o f the

hydrologie source term may be transported colloidally with the remaining dissolved mass

sorbing significantly (literature values o f K f range from ICX) to 1(K),(XX) with a preferred

modeling value of 10,(X)0, Isherwood ( 1977)). Additionally, the solubility o f actinides as

a group maybe limited due to competition to form the same complexes in groundwater.

This would have a significant effect on the magnitude o f the hydrologie source term.

Actinides generally have long half-lives and low threshold values ( 15 pCi/L). Due to

these conditions, all the actinides in the initial source term except for ‘ ‘‘^Cm and "‘‘ ’Pu

have predicted peak concentrations above threshold values at Oasis Valley.

58


59

O f the seven nonactinide radionuclides predicted to be significant at Oasis Valley,

five were modeled as conservative species. Species which have experimental or

theoretical data supporting the assumption o f conservative transport include: ' ‘ 1̂, '"‘C,

*̂C1 and ^Tc (Buddemeir and Hunt 1988, Isherwood 1977, Kerrisk 1985). Although

'^^Eu may sorb strongly, it was modeled as chemically conservative due to the potential

for colloidal transport (Buddemeir and Hunt, 1988).

The results presented show a trend o f the long-lived species as having the most

potential to present a health risk at Oasis Valley. O f the twenty radionuclides predicted

to break their respective threshold values all have half-lives over 70 years except for

'^“Eu, which has a half-life o f 36 years.

Site characteristics which determine the concentration of radionuclides at a

control plane include: (1) initial mass and distribution (2) solubilities (3) rate o f

groundwater flow (4) dispersion in groundwater system (5) sorption unto host matrix (6)

potential for facilitated transport, and (7) radioactive decay during transport. The

quantitation o f these phenomena is critical in providing an accurate prediction o f

radionuclide transport.

Additional assumptions that may significantly affect results include: modeling

radionuclide source as a pulse injection, groundwater flow path and spatial extent o f

aquifer. Due to a lack o f site data quantifying matrix diffuison and nonequilibrium

sorption, these processes were not included in this investigation. Matrix diffusion may

significantly reduce the effective transport velocity o f solutes by being removed from the

higher velocity fractures into pores o f the matrix resulting in lower concentrations at

boundaries. Even considering the conservative method and assumptions use in this


60

thesis, results predict that the summed dose from radionuclides does not break the 4

mrem/yr threshold until 280 years post detonation.

The sensitivity analysis performed identified the mean groundwater velocity and

retardation factor as the most important parameters affecting peak contaminant

concentration and time required to break regulatory concentration threshold values. The

variance o f In K and source size were found to affect the time required to break

regulatory concentration threshold values but peak concentration was not highly sensitive

to these parameters.

Evaluation of Method Selection

This method is computationally fast which was very important due to the number

o f radionuclides in the source term and multiple boundaries. The method also

incorporates the significant physics o f groundwater transport and uncertainty in these

transport parameters. Additionally, the spatial structure of the contaminant plume is

defined. This method however, may overestimate solute concentration when modeling at

longer distances due to neglecting pore scale dispersion.


CHAPTER 8

CONCLUSION

The results of this thesis suggest that twenty radionuclides from the Benham event

may be present in concentrations that are above the EPA human ingestion regulations at

Oasis Valley at some point in the future. This transport modeling has been based on very

scarce, publicly available data and conservative assumptions, as such results may differ

significantly from actual site conditions. Results presented are not intended to provide a

definitive projection o f radionuclide concentrations or migration pathway but an estimate

o f potentially significant radionuclides and to highlight areas which need more detailed

study.

This list o f potentially significant radionuclides could be reduced considerably by

a more rigorous investigation o f the parameters affecting radionuclide transport,

particularly with respect to the actinides. These include quantifying colloid-facilitated

transport (availability o f colloids, sorption o f solutes onto colloids, sorption o f colloids

onto host matrix) and determining an accurate hydrologie source term. Other significant

factors that w ill increase the accuracy o f transport calculations include determining an

accurate groundwater velocity and conducting sorption experiments for the suite of

radionuclides using an aquifer specific matrix.

61


REFERENCES

Andricevic, R. and V. Cvetkovic, 1996, Evaluation o f risk from contaminants migrating by groundwater. Water Resources Research, V. 32 No. 3, p. 611-621.

Andricevic, R. and V. Cvetkovic, 1998, Relative Dispersion for Solute Flux in Aquifers. Journal o f Fluid Mechanics, In Press.

Associated Rocky Mountain Universities and Lawrence Radiological Laboratory, 1965, Proceedings from the Symposium on the Peaceful Uses o f Nuclear Explosions, April 26-29.

Beilin, A., P. Salidin, & A. Rinaldo, 1992, Simulation o f dispersion in heterogeneous porous formations: Statistics, first-order theories, convergence of computations. Water Resources Research, 28, 2211-2227.

Blankennagel R.K., and Weir, J.E., 1973, Geohydrology o f the eastern part o f Pahute Mesa, Nevada Test Site, Nye County, Nevada: U.S. Geological Survey Professional Paper 712-B.

Borg, I.Y., R. Stone, H.B. Levy, and L.D. Ramsport, 1976, Information Pertinent to the Migration of Radionuclides in Ground Water at the Nevada Test Site, Part 1: Review and Analysis o f Existing Information, UCRL-52078 Part I . Lawrence Livermoore National Laboratory.

Brikowski T. And G. Mahin, 1993, A Conceptual Model and Preliminary Estimate ofPotential Tritium Migration From the Benham (U-20c) Site, Pahute Mesa, Nevada Test Site, DOE/NV/10845-25, UC-703, DRI Pub. 45109, August.

Buddemeir R.W. and J R. Hunt, 1988. Transport o f colloidal contaminants in groundwater: radionuclide migration at the Nevada Test Site. Applied Geochemistry. V 3 p535-548.

40 CFR 141, U.S. Environmental Protection Agency, “ Ntational Primary Drinking Water Regulations,” Code o f Federal Regulations, Washington D C., 1995.

Cvetkovic, V.D. and G. Dagan, 1994, Transport o f kinetically sorbing solute by steadyrandom velocity in heterogeneous porous formations. Journal o f Fluid Mechanics. 265,189-215.

62


63

Dagan, G., 1988, Time-dependent macrodispersion for solute transport in anisotropic heterogeneous aquifers. Water Resources Research, 24, no.9: 1491-1500.

Dagan, G., V. Cvetkovic, and A. Sharpiro, 1992, A solute flux approach to transport inheterogeneous formations 1. The general approach. Water Resources Research, V. 28, No. 5p 1369-1376.

Daniels, J.I., R. Andricevic, L.R. Anspaugh and R.L. Jacobson, 1993, Risk-basedscreening analysis o f ground water contaminated by radionuclides introduced at the Nevada Test Site (NTS), Lawerence Livermoore National Laboratory Report, UCRL-ID-112789.

DOE, 1994, United States Nuclear Tests, July 1945 through September 1992, DOE/NV- 209 (Rev. 14), U.S. Department o f Energy, Washington, D C.

DOE-NV, 1996, Preliminary Underground Test Area (UGTA) Regional Flow andTransport Model and Supporting Documentation, U.S. Department o f Energy, Nevada Operations Office, Las Vegas, NV.

DOE-NV, 1997. CAB presentation, Beatty, NV

Ekerman, K.F., A. B. Wolbarst, and A. C. B. Richardson, Limiting Values ofRadionuclide Intake and A ir Concentration and Dose Conversion Factors for Inhalation, Submersion and Ingestion, Federal Guidance Report NO. 11, U.S. Environmental Protection Agency, EPA-520/1-88-20, Office o f Radiation Programs, Washington, DC, September 1988.

Freeze, R. Allen, 1975, A stochastic-conceptual analysis o f one-dimensional groundwater flow in a nonuniform homogeneous media. Water Resources Research l l . n o 5:725-741.

Grindrod, P., 1993, The impact o f colloids on the migration and dispersal ofradionuclides within fractured rock. Journal o f Contaminant Hydrology, V 13, pl67-181.

Hechanova, A.E., 1997, A Methodology for Estimating Residual Radionuclides from Underground Nuclear Device Testing in Nevada (Draft), Nevada Risk Assessment/Management Program, HRC, UNLV, Las Vegas, NV.

Hoeksema, R.J. and P.K, Kitanidis, 1985, Analysis o f the Spatial Structure o f Properties o f Selected Aquifers. Water Resources Research, V 21, p563-572.

Isherwood, D., 1977, Preliminary report on retardation factors and radionuclidemigration, Lawerence Livermoore National Laboratory, UCID-17551,22p.

IT Corporation, 1997, Value o f Information Analysis for Corrective Action Unit No. 98: Frenchman Flat, lTLV/10972— 192, UC-700.


64

ICRP-30, Limits for Intakes o f Radionuclides by Workers, Publication 30, Parts 1,2,3,4 and supplements. International Commission on Radiological Protection,Pergamon Press, New York, NY, July 1978.

Kerrisk, J.F., An Assessment o f the Important Radionuclides in Nuclear Waste, Los Alamos National Laboratory report L A -10414-MS (October 1985).

Kerrisk, J.F., Solubility Limits on Radionuclide Dissolution At a Yucca MountainRepository, Los Alamos National Laboratory report LA-9995-MS (May 1984).

Knapp, R.B. 1989, Spatial and temporal Scales o f local equilibrium in dynamic fluid-rock systems. Geochimica Cosochimica Acta 53, 1955-1964.

Laczniak, R.J., J. C. Cole, D A. Sawyer, and D.A. Trudeau, 1996, Summary ofHydrogeologic Controls on Ground-Water Flow at the Nevada Test Site, Nye County, Nevada, Water Resources Investigations Report 96-4109, U.S.Geological Survey, Denver, CO.

McDowell-Boyer L.M., J R. Hunt, and N. Sitar, 1986. Particle transport through porous media. Water Resources Research 22, 1901-1921.

NRAMP, 1996, Preliminary Risk Assessment - DOE Sites in Nevada, Volume I and II, Nevada Risk Assessment/Management Program, Las Vegas, NV.

Ogard, A.E. and K Wolfsberg, W.R. Daniels, J. Kerrisk, R.S. Rundberg, and K.W. Thomas, Retardation o f Radionuclides by Rock Units along the Path to the Accessible Environment, Materials Resource Society Symposium V 26, 1984.

Ptacek, C.J. and R.W. Gillham, 1992, Laboratory and field measurements o f nonequilibrium transport in the Borden Aquifer, Ontario, Canada. Journal o f Contaminant Hydrology 10, 119-158.

Shapiro, A.M. and V.D. Cvetkovic, 1988, Stochastic analysis o f solute arrival time in heterogeneous porous media. Water Resources Research, 24, 17I I - 1718.

Sheppard M l. and D.H. Thibault, 1990, Default Soil/Liquid Partition Coefficients, Kds, for Four Major Soil Types: A Compendium. Health Physics Vol. 59, No 4,471- 482.

Smith, D.K., 1993, A Review o f Literature Pertaining to the Leaching and Sorption of Radionuclides Associated with Nuclear Explosive Melt Glasses, Lawerence Livermoore National Laboratory., UCRL-JC-113370.

Smith D.K., B.K Esser and J.L. Thompson, 1995, Uncertainties Associated with the Definition o f a Hydrologie Source Term for the Nevada Test Site. UCRL-ID- 120322, Lawrence Livermoore National Laboratory.


65

Rubin, Y., and G., Dagan, 1992, A note on head an velocity covariances in three- dimensional flow through heterongeneous anisotropic porous media. Water Resources Research 28(5), 1463-1470.

Taylor, G.I., 1921, Diffusion by continuous movements. Proceeding London Mathematical Society. 20, 196-212.

Thompson, J.L, A.B. Kersting and D.L. Finnegan, 1997, Plutonium in Groundwater atthe NTS: Observations at ER-20-5, American Chemical Society Symposium, Las Vegas, NV.


VITA

Graduate College University o f Nevada, Las Vegas

Lloyd Thomas Desotell

Home Address:5802 Anniversary Street Las Vegas, NV 89119

Degrees:Bachelor o f Arts, Environmental Studies, 1992 University o f Nevada, Las Vegas

Thesis Title: Radionuclide Dose Assessment o f Groundwater Migrating from theBenham Event, NTS; a Solute Flux Approach

Thesis Examination Committee:Chairperson, Dr. Roko Andricevic, Ph.D.Committee Member, Dr. David Kreamer, Ph.D. Committee Member, Dr. Klaus Stentzenbach, Ph.D. Graduate Faculty Representative, Dr. Mark Rudin, Ph.D.

66


" 'A ,

V.v°

IMAGE EVALUATION TEST TARGET (Q A -3 )

'/.■

/ ,

e

1.0Li |2£|J0 liàla Lmc L6 tià

1.1ë L& 16L. .

N jâu i | . 4

150mm

4%:

<9

&/ .1y

/A PPLIED A IIS/MGE . Inc1653 East Main Street Rochester. NY 14609 USA Phone: 716/482-0300

- = ' - = Fax: 716/288-5989

0 1993. Applied Image. Inc.. AS Rigtits Reserved

j f V

\


Radionuclide dose assessment of groundwater migrating from ...

Documents

Transcript of Radionuclide dose assessment of groundwater migrating from ...