K(f/A (J IA' Y I> 1/2/

SVerslon of. February I

.{::, : - Ear th. and.-,Space Sc ie n ce•s , :Di vis~ion,:•i:•.: -is. ::::.,i

.. _ . - ' . -..';> i ::)

-. '. .. . . . .

Los .Aia os, e . Mexico 87.545

- .. .. -(.. . .: Mi e- o e 3 5- Leve1." : -."I) '> ,7.

p.'

9307090240 930617 PDR WASTE WM-II PDR

m

.LIST OFTABLES

- -.::i.::111.-.:I . .: . GEOCHEMICAL/GEOPH SIC L " ~ ~ ~ A " -: -::-. .,Str ti

. ,..>:C o.,y' LS OP-U E :t

,ۥ- I EXEUIE Sou I*(ATR7

IV. NUMERICAL SIMULATIONS A. TRACR3D j B.1 Scenarios.

V. RESULTS." A. Uranium Transport B. Technetium. Transp-ort -C. E f ecft 'of Sorptionj D). Effects of Flow

-VI. DISCUSSION. .

A... Ceocherical/CGeophysic B. Transport.Calculto

a laio

* - 9., . 'S 9 * S

* ... 9 9 5 9

* . 9. 9 . 9. 9

* . S 9 9 .9 *

*.........

. *... *�. .14

*..................1� * 1:

0

2 3

14

* . : Y . 1 6

*: 16

* 17 20

* 21

21 22

VII. CONCLUSIONS AND RECOMMENDAT A. Concluslons .. B. Recommenidations.

REFERENCES. - .

a

3 -

5 67

. 9

C-

16

21

2324

*214

. . 26

3

-::•:,' :,:: .:

15

cv!

C

LIST OF TABLES TablI.rtes of the Stratigraphic Units ', ;.

: Tabl e+A: Avarlfi+an: Est-Imae, Distribution Coe3,ioients, Kd's, r

Table I•I I ,-or t he. Numer icall ...Simulations I of Uranium and Technetium ,raspr

'LIST OF FIGURES:

Figure 1.. •i -rpresentation of Yucca Mountain stratg scale)., rpy.ntt

S Figure2. Liquidphas:e concentrations during transport of 2)1U1 through un-. "saturated'.¥Yucd•cWouta-n' tuff at normal flow rate. Adjacent lines differ in -oficetrtfban[•+b a factor+ of 10. u .: .5.mn/yr. t,- 5,000 10000. and '15,o000 y 1-(Case•1) .- I

Figure 3. Sorb•edd'cncentrattons. dbr'ing transport o 2 ,U through unsaturated acca Mountain tuff at normal'flow rate. Adjacent linesidiffer " in, concentrAation•by, a factor of 10. u 0.5 mm/yr. t 5,000, .10,000 and 15..(C. se 1)

* Figure !'4 L.~qiu-~n cnentrations during transport of 23Uthogun ' saturatedYuca-.'unta-n tuff at high flow. rate. Adjacent. lines differ

in, cncentratioxi.by•a•y factor of 10' u 4.5 m/yr, t 5,000, i0,000, and I .OWse 2)

Figure 5. Sirbe concentrations during transport of 23 1U through unsaturated iiucea-Kuntain tuff at high flow rate. Adjacent lines differ in# concentriatt.6by a'. factor of 10. u 21.5 num/yr. t = 5,000, 10,000, and 15,000j 3(+Case: 2) Z

Figure 6. Co5pusuif'of liquid-phase concentrations of 238U after transport for.10,000 jr'.al 4 fferent flow rates (Cases I and 2). Adjacent lines differ ini -o.iifii:tion by a factor of 10.

Figure 7. .Liquio-pnase concentrations during transport of STc with sorption throug)•iUnsatuprated Yucca Mountain tuff at normal flow rate. Adjacent lineis ditfer.1in concentration by a factor of 10. u 0.5 mm/yr. t-:.5:•0K_10,000, and 15,000 yr. (Case 3)

Figure 8. Sorbea.iOncentrations during transport of "Tc with sorption throughu ... turated Yucca Mountain tuff at normal flow rate. Adjacent lines-.differý!-_ concentration by a factor of 10. u = 0.5 mm/yr t z 5,000, 10 ,000 and 15,000 yr. (Case 3) .5 " /"

Figure 9.. LlquiA• &se concentrations during transport of "Tc with sorption through un aturated Yucca Mountain tuff at high flow rate. Adjacent lines differ in ' oncentration by a factor of 10. u =45 mm/yr. t : 5,000, .10,0001 •and 15,000 yr. (Case 4) Z " y

3

• • ,. . • . . .. .

r e -- ....... . ...

fi~ure10.Sorbed concentrations during transport of t T~ihsrto hrugh unsat urated4Thtouta in' tuift at h'ih flw ate. Adjacent

lIn. concentrati by a - facto oi 0 4.5 r/yr. "-2 5,000, 1OO00, and 15,000 yr•. K(ase-ii4}q->-

t1r gure I. Comparison of liquid-phase concent ocfarfter and ransport hsorption fr di ,er ertf rate-(ases 3

",,n .. :4)., Adinaesd e lines di ffer In•oncentration b.a factor of 10.

S.ji$:•Fig iti-12. a Liquid-phase concentratiosn'trarnsport ofte'!:To without

.14so •QSrptiob through unsaturated Yucca- Montain tuff, at normal Cflo rate. - '-•/:" "!:6)..ddcet linesndlbtyii f c naentra actor of 10. u 0.5

:::!:it-! :AdJacent lines differ. in concentrationi by a factpr of.10. u2 O.5

, J:!."ig'4,r.•••e:6. Comparif.son ptof .liuinhae concentrati onse rtof To£ afteratr" :.

/yr., t 5,000 an i1hOdr. (-15 e 6 6) yr. (Ca '

~~~~~~ and 6).in Adaetlnsdfe ncne tratin by a fact of 10.t

~-~ij~iute15. Efeto opino iquid-phase concentratio of Tpor afe

ttansport through Yucca, Mountain tuft at normal flow rate t for 10,000 yr.

A d J Adjaeent llnes differ in concentration by a factor of 10. u :i 0.5

0(Cases 3and 5)

ctCFltze 6.Efec ofsoptononliui-phs'concentrations of9T f~tcr

I .< ,transport, wth.og-hc Mountain 0 tuf at nra lwrt o 000'r and6).Adjacent lines differ in co'ncentration by a factor of 10 a 0.5.

• .t- gure 17. Hypothesized model of the flow regime through the hydrogeologic untsatp6t Yucca Mountain.

is. I r 6'Efc fsrpin nlSi-hs-cocnrto fT fe

rasprttruhuc.onantffa o. fo7aeo 00 yz.

Adaet'Ss ifr'icnetainb a-atro 0 '--S

4

•GEOCHEMISTRY S SI'MATO -OF YUCCA MOUNTAIN. MODELING .THE TRANSPO(t46F URANIUM AND TECHNETIUM TH)ROUcH."TH E~$5;~ TUDTFFS

.. ".. G A. Ceder er-/ý ", Eriec-(.r.eenwade- and Bryan J. Travis Earth~ and Space:ýScience, Division

ASSTRA4CT ... ..- .... thi.report Preiminy beline calculations for the transport of"uranium and technetmthrgh the unsaturated. zone at " "Yucca. Mountain,.- Nevada, .are'-pre-ented. Uranium is representative "":ofthose radionuclides ith extr.eme .Ylong half-lives and-high al..ues 'of -Sorption coefficientsg.',,

of "V 'e r - f .i iecnetium is representative of ... " theost, solubeand.fastest movng 'radionuclides. First. a founat on for the. alculat.on is iscussed... This foundati* a "refere-nce d eache.o ialgeOphym model, contains thec CD. i i;;. - .... stratig.raphic- petrologic, •hydr"o"g'eo"'ogc ,-""geochemical, andematerial property data for the..Yucca .Mountai. site. Second, the Integrated transporti of uranium-.and technet 'rom the reposltory to the water table is modeled. . Aflexie.e case flow scenariO and an extreme-case, flow sc Y : s'(c rii h .e h e iu b ao p nd the exre s fo scenario: ar e,. >•n the transport calculations Becauseý of theuncertai ntyassociated-with technetiums sorption-of technetlui is neluded an": then neglected In both flow scenarios. Thus, six transport!.aseare

modeled. Re that the estimated uranium tran.po ?is only slightly sensitive to the magnitude of the flow because, orption has a significant effect on the. retardatio..The .re.sults ,:r technetium transport closely Sresemble the resul~ts for uranumý:teAnsport.;

The flow dominates the transport when sorptIon Isnegle •Although the quantitative results using this prelimtinarymodJ.'. Indicate that EPA requirements for a nuclear waste repository.Wil1- beimet easily, the model must be updated to a more realistic form before this conclusion can be regarded as dependable. Thesee.preliminary baseline calculations will be used as a basis to Investigit ,'i-the effects of physical and geochemical processes on the long" termtransport of radionuclides at Yucca Mountain. . ,

. : ..;• 'V;::.

------------------------ ----

5,

I. EXECUTIVE $U5 MARY'r

An area.unsaturated fractured turfs at Yucca Mountain, Ai;ýdotent si 3te~sbei ng considere t or gelog:torag,

.f.- igtfh_ vel i e waste The Environmental Protection Agency (EPA) 1- 0CFI ............. •lmits the cumulative releases of ma.y.radionuclIdes

f-rom t o tohe accessibleenvironment for 10,000 yr after dis

po~sal' (U S 'EPA 19 5)-.'-An estimate. of the, transport and retardation Of radionucUdes Is ýnecessary to'assessthe expected postclosure performance of: a potential, reposItorj

To, " - th":'- need-a:continuing series of, calculations is being. carrieh adionuclide transport from the repository to the

accessible environ-ent using; the late.r. available- models and data.- :.These, c"lculations w-illý be eto investigate the effects Of physical and' geoc-heical. processe o,.n' the long-term tran~sport of radionuclides at Yucca--, Ycu ina -This rieport presents the results of a first set of calculations cf radionuclidet.rans'port" from the disturbed zone to the water table. The transport pathway-,43 sthe unsaturated zone. The geochemical/geophysical. =,=,del used iiathes alcu-lations is a preliminary one; much more information

. . "is needed before .-e ge'ochemical/geophysical model can be regarded as suffiSc :ently -ope ý6 6 ipeitos t co e tt it can be used for making dependable predictions.

"The.transportioýitwo ýradionuclides is modeled: 2"U and-._-ý Te Uranium iaradinuclides withlong half-lives and high values of

sorption. _and .technetium is representative of the most soluble and. fastest movin dionuclides. Two flow velocities through the tuff are used for each radin oCi4'1de, and for each sorption condition: an expected

a~ velocity.:.OS • •.mm/yr and an extreme average velocity of 4.5 mm/yr. Thus, sixi transport s4ennarios are modeled.

The following. o'riconclusions were made from these simulations: 1) For ur ,the estimated transport is only moderately sensitive

to the magnitude f, the flow. The insensitivity is caused by the sorptive propert esof the tuffs of Yucca Mountain. (2) For: tec i the sorption distribution coefficient is only slightl y less tUjaji:that for uranium in the repository unit, and the results for technetium transport closely resemble the results for uranium transport'.

6

(3) .Flow dominates,,the.'transport, when sorption is negligible. (4l) 'The uranium, ba~se~d:-`on the prel1ini ry ecela/ep ia

.. : -- repsit•ory is: locate4in•: th the expecte •and"extreme flow scenaios•.•i• !-i•• (5-)~~~~~~~~ Th.tcntu~bsdo h rlnary,,geochemical/geophysical"

model, aendpe liiantransported benyonhd the

are mad tobe ter de erm ne the slg lfl anep s t r a -d im o t n e o -p oe-s -i.. ..

af trbtteg "transphi unit in, hiha thew " .ocated.in both :he • 1- . . . ._.... . . . the .expected and extree.eem ee lylo..nsten r s.

moex ested o pothlue waerfotable- und £•ter. c- o endtions reof g ltowrate.-.::.-:•i,..

"bed o t(b)he preldaiary gee. geohemleal/geophysioi wthl

theucc Mountde ine cant foe rearndcd ar satsateoryana .sial base.

The for prelow imnare erations. or1futone stuesa data ahqui'beyondith.o tra-iority. c unt: Iwhcterpo' y13-. loae Inbo te,..

aemade3,to btter determiabe thder cignificancead no rtan lof protes. e

a e2ting tranlorth e amout h rafternzet ie morsensomtpe to oelr and tao b aess

(s1)pd or datao a are needed beforeo the seorheic gepypcal model. isus thin

Te.for lwn prdciecsimulnations. Collection ofthdesean datashudb aqu high~

of Yucca Mountain, especially the stratigraphic units dlreetl y underly

ing the propoeosie repository. (3) Additional simulations should be carried out coupled to mre com

plex models of the flow paths and stratigraphy within Yucca Mountain.

11I. ]NTRODUCTION The Nevada Nuclear Waste Storage Invest itieons NNWSI) Project is

charged with studying the feasibility of placing a high-level nuclear waste repository in the voleanictuffs beneath Yucca Mountain,t Nevada , The mined geologic disposal system must meet the sysoem performance objectives for radionuclide releases to the accessible environment as required by the EPA

7

20 CFR Part 191 (U.S. EPA: 1985). To assess the expected postclosure per

formance of a potential 'repository, radionuclide transport and-retardationý'c

estimates -obtaine~d from numerical s imulations are ,nec essar y. ,-One ongo .Ing,

investigation In the .NNWS1i Project is to determine"'thesi ignificance and

relative im~portance of, the 1hscladgohmcl proces'sesa4ffecting ra'dionulide transport.- As one part- of, this sp'c fic..•vestigatio ', the

transport, of uranium and.technetium through the-unsaturated"zone has .been

modeled. The calculatlon3peetdI ths report. represent -as smplesto

"case scenario, for radionuei de".transport. These prel imnary baseline cal•

culations will.serveas.reference points for comparison with'more complex,..

integrated.transport calculations carried out as site characterization

proceeds.: This report describes the modeling of the transport:of uranium 'and'

technetium from.the repository to the water table..Uranium-is

representative of those radionuclides with extre mely"long half-lives and

high values of- sorption, coefficients (Kerrisk 1.985}.--Technetium is*:

representative of the most soluble and fastest moving radionuclides.(Kerrisk .

1985). Estimates of the transport were made under twotflow condit'ions for

both radionuclides and two sorption conditions for. technetium. A total of

six scenarios' were modeled. These calculations were made using-the computer

code TRACR3D (Travis. 1984). Input for the computer code comes from a com

prehensive, referenced geochemical/geophysical model (Creenwade and

Cederberg 1987); this model is discussed in Section III. The

geochemical/geophysical model contains the current stratigraphic,

petrologic, hydrogeologic, geochemical, and material property data for the

Yucca Mountain site. Known repository data and estimated values for data

that are unavailable are given. TRACR3D and the scenarios that were modeled

are described in Section IV. Because of the unavailability of data and an

incomplete understanding of all the processes involved, two flow scenarios

are included: a possible- expected vertical flow rate of 0.5 mm/yr and a

potential extreme flow rate of 4.5 mm/yr (Montazer and Wilson 1984; Wilson

1985). Also included is an additional set of calculations in which zero

sorption of isTc is assumed for both flow scenar-ios. An extreme-case

scenario for radionuclide transport from the repository to the water table

occurs when the higher flow rate of 4.5 mm/yr and zero sorption of 9Te are

combined in a single simulation. The results of these simulations are

8

9

-presented in Section V. A discussion of the results. is9 givennSeto i Conclsionsfrom his sudy and. recowmiendat Ions for futuesuisaddt * acquitonae given~in Sgetion V1T .

11.GEOCHEmICAL./GEOPHYSI CAL MODEL.. .

Befrean ste-peifc adonuclI de transport is modeled, the., current S stratigraphic, petrologic,:hydrogeo-logic,, geochemlcJadzaeiidt o tha s te ho ld be ol ect d. Th s s ct on summarizes. the da~ta that were repor~ted, in a, comprehensive,- referenced geoc~hemical/geophysica! model (Greenwade and.Cederberg 1987). The ýknown repository. data: pertinent to the

Yucc Montai sie were compiled, and unknown parametervauswreti mated ased n available da taý. As si#-_a character12atio'poed,'hscn

ceptual geochemica'l/geophysj0Ica1 .iod e 1, will1 be continually updated and.

revised to reflect the currently gathered' in-formation: and data..24 A. Stratig-raphy

Astra~tigraphic model , is used- to divide the tot l t a s o t p t i e , e ~ .tltrnprtOt from the repository to the water tabl'e).into geologicalditntnts

The stratigraphy of the Yucca Mountain turfts is quite. complex.Igera,< there are alternating layers of wede ndnowldd ufs Tegeloi,. hydrogeologic, 7$eecemclpoprisvrycnierbyaog units.-C

t h e t u t s a d d f e e c s i a e ri a ic m ps t o n T hsvvaa tor ya C- ignficntl afec th ov ra l e tim ted tr nsport. y-mo C,.-i F i g u rer1a t hon s i nra s c h e m a ti. fd ut heer a s t a t g r p h o Yu c c Mountin (reenade ad Ceerbe g r en s 197) The dta egreph e mode. gve i F i g . 1u f i s l oa t ed mi d w a y b e t wncel l U S w e l i n a d U W G 4 i Y u c M ou n a n (Fg.5;frm rtz t l.19n. mathris loaostion. waseselce beciausema signficntr reative erro oestimates asoiated wthanpotheeeains.r ie (71b Frligur theovlus forsceaicfr the elvton ebaned rom dtraillh hole data (Cunainel 1986).ad argnd iserr a 97) esalThea srtechnique usedet aclt

data (Mtherocae 1963 ay b97)etheen welevtos use -5ad SG- in thica reoruwre taken (From 5he fro Oataz Baet aTlF 1984a).bcas thes dattinas bselectvery beausyet acces, renatise caero etmthe T 9 Dat Basscaed proides the saelevaluens wore thven elevatgionsa tho ause frteeeain obtained byo drriln (apel18) Theeoelevdatio

13is defined to be at the static water level (SWL) 1 i and the bottom potential repository slab is at an elevation'of 257M9169.

FV For the unsaturated zone' the 1ua broken in6 ht a-->'"-i?::i.u distinct.therm 1/' units,, each Wit th a mechanical propert ies:(Ortz et i -> i -•:(1)). TCw, Tiva- Canyon welded; (2)-Tn, Paintbrush flonwelded; (3) T a-i< Spring welded; (4) CHin, Calico Hills nonwelded; (5) PPw.,' Prow" Pas

~~ a- and (6) CFI~n,.Upper Crater Flat nonwelded. TheTw itsdvi subunits,. TSw1, TSw2, and TSw3. -These three subunits have distin

* a'- i:a and sorptive properties (see Table Ii). The CHn-.unit• is divided -subunits. The flrst subunit, CHnlv, is a vitric layer.. CHnlz is

--.layer. CHn2 and CHn3 are both assumed to be zeolitie layers havi -. - mAterial- properties.

"~-: B. Properties of the Geologic Media..

The major properties and characteristics of the geologic med : }.: affect transport are saturation, porosity,.and dispersivity.,-The.

, . bulk density also affects transport but only-through, the geochemi, process and the particular definition of equilibrium sorption usec

- *-~-,-a -report. -Values for saturation, porosity, and matrix bulk density of the subunits are given in Table 1. The values listed in Table

-... , ... a-•a-a- unweighted means of drill hole data provided by the TUFF.Data Base S...i:• :..ii!:-:•: !i!!i 198.,g 6b, TUFF 1986c). t * :-.: "."a .a-a-The dispersivity is often used to characterize the dispersion

...-; . mixing and spreading of the plume caused by microscopic velocity v -2~a-

ve:oeith

w.ithin the pores (Bear 1972). In recent years, studies have sugge 'dispersivity is not constant but rather depends on (1) the scale o - heterogeneities (e.g., fracture spacing or the spatial variability

"hydraulic conductivity), (2) large-scale heterogeneities (e.g., di geologic units), and (3) the mean travel distance and/or scale of (Matheron and de Marslly 1980; Pickens and Grisak 1981; Gelhar and

'a. . 1983). a"a--"-- _ _ For the Yucca Mountain site, dispersivity data and general in: . a-: ':. concerning dispersion do rot exist at this time. Minimum values f(

- gitudinal and transverse dispersivity were chosen based on fracture "a--a --. " (Hontazer and Wilson 1984), a scale of local heterogeneities. The

of the

ilow1ng-:

al, 1_984):

Sw, Topopah

3. welded;_.-,

red in three.

act material

into four

a zeolitic

ng distinct

ia that

matrix

stry

d in this

for each

I are the -(TUFF

, or

ariations

sted that

f local

of

stinct

the system

Axness

formation

or lon

e spacing

effects

10

I

of flow and sorption. are beinginvestigated in -this report, therefore,- mini. m•. diSpersiVity values were chosen so that the effects of dispersion would

be ýnegligible. ,uvle olngitudina~l dispersivity, Th-Lvle'o was set to 10

c-, and the value for transverse dispersivity,' was set to 1 cm. For.

"these.prelimineary cAlculations, each.ýstratigr'ap.•hcunit was assigned the

same values of longitudinal-and transverse. dispersivity.

C. Hydro~eologY

Fluid -flow at •Yucca:Mountain- occurs. through heterogeneous, anisotropic,

fractured tuft.. Little is known about the natural groundwater: flow in the

unsaturated zone, and investigations in this area are only in the prelimi

nary stages (Xontazer and Wilson 1984; Roulon et al. 1986). Analyses

indicate that the distribution of the vertIcal percolation. is nonuniform in.

the unsaturated zone: (Montazer and Wilson 1984). .Preliminary calculations

(Roulon et al. 1986) indicate that because.:f the dip of the-stratigraphic. .

units and because of their hydraulic properties, a significant proportion of

the flow above and/or below the proposed repository horizon may be diverted

laterally into a permeable fault zone. The magnitude and location of the.

calculated lateral flow depend upon whether matrix-flow or fracture-flow'

conditions are assumed for the highly fractured units, upon the flux*

specified at the g-round surface, and upon- the hydraulic properties assigned

to the fault zone. The results of these calculations are controlled by

poorly known hydraulic parameters such as the characteristic curves (Roulon

et al. 1986).

Although the stratigraphic analyses indicate the presence of tipped

beds and early hydrologic studies indicate that the tipping may affect fluid

flow, too little is known about these effects to incorporate them into our

model at the present time. In this report, it is assumed that the recharge

rate is applied as a constant vertical velocity field over the entire

mountain. A simplest case scenario for simulating radionuclide transport

occurs when a constant vertical velocity field is assumed. These prelimi

nary calculations will serve as reference points for comparison with the

more complex, integrated transport calculations carried out as site charac

terization proceeds and the hydrogeology is better understood. For all the

stratigraphic units, the flow is presumed to be matrix dominated with

insignificant lateral or fracture flow. An expected value for uz, the

11

average verticaI. flow rate, 1',* 0.5 J'ranyr, whereaa~4em vaemybea

a .... -xteme value may be as

_ 'hhyar •(•_r m(ontazer and Wilson. 18lI- ..I..

V ~~~D. Ceoch emist Q.- %1L$* .

The, main echemlcal proces esaffec~1gt rasotaesrtoln g.-eoj!!i,.!. sesee ng_ tXZP.T13p:'t•-a sorption and•zoleculaP diffusion. One pa~rarnter that can',be assoiatd wit h qi6r ...sorption, J i the distribution" coefricient, J(c.-a-the parameters asso c Iate d 'with molecular dif'fusion are the dif~fusIonT-coertjicient D0

Sand the constrictivi- .. *Measure O~f the partitioning et ofe ewe t~sldadaqeu..h

~~s .... between,;•: in aqueous-phase under .ccrnditions of local equil brum•- The fcil n uat gives the relationship between the solid ~and aqueous-phas ~ciz~osfrteds ~~~~~~~~~~s concentr.. .:.:<• •. : ..... :-..•ations for the distribution coefficient sorption mdel

. sorbd -phase concentration, (g/. d -equilibrium distribution coefi

C aqeousphas conentraticn, (/m

sThe larger the val*of:K the-morelsolute is so-b ed• o'o the solid phase. Forý both aqueous- and. solid-phase concentratins ig/ta cm offlid ~~~~~~n , n, to alC(ot ) of fluid,, . :. .- ::•:.

Molecular •diffusion causes mixing o the:"ontamin int3 he"ld of te:.cnt• '•"ithe rluld beacause each solute has its own pathline-with rdomainespacs, I t~l ow domain as well as its own velocity along that pathline.- The,.tra -h~p tim. ewe S... • ..•ime between

points of interest 'i 11 be lengthened I f the flow v relt ively small so. that enough time exists for a significan".. 1of the radionuclide to diffuse into the fluid in the surrou' te P9matrix materil.! In this report, constrictivity is considered to beacoobiation of two factors: the measure of the deviation of the pathlne otansportfo hline6f-trnspor froma straight- path (the standard definition of tortuosity), adthe narrowing of the pores, usually called constrictivity. For most real rous systems, it is. impossible to separate the two factors (Satterfield 1970) it is also this combined quantity that is determined experimentally.,,,*,,,

12

13

Cy.

CV

Experime"tally. dtermined (Thomas 1986) and extrapolated values ,"J ; " :L(Creenwade-and

* aregienwd Cn..ederberg 1987) for apparent distribution coefficients' 'a, o theradionuelides "To and ;IO. They. re called apparent disru on ericlents to acknowledge the fact that-the reaction :"•i)•!:::"in•: the : 1-- have r-eached a state of reversible equiljrj : (...- as ge:.1986) [ableIi, values based on the experimental Kdaa as.. ign. d to.he riate thermal/mechanical unit by correlating' te depths Of the Smes ( 3oa 1986) with the stratigraphic information .(Ortiz, et alý 1984:) A a first.approximation, the unweighted mean was ; .t a k e n :O if - _a l l trh t h p a r tg h t m eaaw a tae of-a edatafor- a particulr unit regardless of the drill hole :from which.-the Sal had:°originated. Where data were unavailable,= distr..Ibu.tion, coef f.icient data, were assigr..d, to stratigraphic units :based on the unit 'Is.mineral c•m•Osition (Thomas 1986) and material type (Broxton. 1985;, :1986). Details are given in a recent report (Greenwade and Cederberg 1987). svrateigrathicn s " t characterize themineral compositions of the -. S.r...g"p:hiciunj~ are being done at Los Alamos (Broxton 186. A va tfon • in Table i.• • • •.-.• - .. . . . .1. 986. ) An -obser-; v.. ti.n in Tabl .. -.. i, hatthe distribution coefficient for techneti, is non-zero In sievera •. tratigraphi, units. Thi. is important because it is. usually :... as u e,,t:::; `-•-'t. U" • t e n b s osu . a�d ttchnetium exhibits little sorption, much less than observed. experimentally in the TSw2 unit (Tien eL al 1985) " Val u es .f M tL- ll.ecul a i

ni.u .were nor ..a a eulcab r diffusivity coefficients for uranium and technetium were. not'." l'avhbl'e"from the results of current Los Alamos investigatbons':(R "-rg 1986) or from the literature. Therefore, D is set' equal to 1..5 .X0 .•1.0P based on an expected value of 1.46 x l0o cm2 /s for an element suc as1 technetium (Rundberg 1986). This is consistent with values for-other mateials, .as diffusion coefficient, for substances diffusing through bulk at.er.are normally around 10- cm2/s (Reid et al. 1977). Values for the°.contric'ti-, vity coefficients were obtained from the results of diffusion cell exper••i.nts (Rundberg 1986). For the welded units, Y 0.037, and for-the.onwelded units, 'r. 0.030. c

E. Source Term

Another-area of active investigation in the NNWSI Project is estimating a source term for the'radionuclldes based on factors such as waste package design and volume ýof-water intercepting the waste package. This work is principally being done" at Lawrence Livermore National Laboratory (Oversby

J i

M7-

a (•ODC) + V*(6uCy}= V.(oGT DPVC)

+ V'[EoDV(PC)] - CXOPC

- PP mKd(tC + XC), (2)

where

.1986). Preliminary estimate3 Of the, ource--termnsor technetium and.' ranium

are. Co Tel- : 2.6 _X 10* gcmu (Oversby and Wilson 1986)_ -and. C0 "U

5.0 x, 10' g/cm. (Oversby: i986i.., Z-n- thd t~ransport., Iclcu ationS' thatý were,ý ý carried out for this reprt_,.th value given above eusedas consa

concentration sources 'o fthe rad i onu'cl des

A. T•ACR3D .

The TRACR3D code (Travis 1984~) was used to perform the -transport cal

culations discussed-in Section V. TRACR3D solves _the, equations. of tran•s.ent. two-phase flow and multi-component, transport in deformable,* heterogeneous,

.reactiveporo~us/fractured media. The code uses.an implicit :finite--.

d! ff e rence technique-to solve the flow equations 3And asemi-I mplicait'finite

differe nce method to solve the transport equations. - One-,4 two-, or three -,

d Mens Ionai grids3can be used in either Cartesian or cyindrical:

coordinates. Verification of the TRACR3D transport model was carried out by - "

,c omparing. TRACR3D With fi've analytic-. olut ions to. problems .involving tracer,

transport (Travis. 1984:).-, It was shoi.ljn that TRACR3fl..can accurately solve the model-'equations. Validatlon of the TRACR3D transport mrod elwa .

demonstrated by comparing simulation results with the results'of several

experiments. Two experiments that have application to transport at:,Yucca.

Mountain are (1) di-ffusIon of a sorbing* tracer from a well-stirred solution.

into a thinwafer of tuff and (2) migration of radioactive tracers from an

underground nuclear test to a nearby well-as a result of pumping in the

well. The model and experimental results were int very good agreement.

(Travis 1984). .

For the simulations presented in this report, the equation of mass

transport for each radionuclide is given by

I. 3 partial derivatj th re.pect to time;

""- sattrationf':v( uid/volue of void.;

f Itlud den3 ty , vi iý.Zb -U - average por wat"evekc ity vector, cm/S; C ..radionu ide con-cetration, g radionuclidelg fluid;"

* I co3 cosrictivityý,, dizensionleS3;

" os e molecularddiufusvity, o radionuclide, cm AS;

" hydrodynamic,' ispedsio d ten'SO ane i so;

X . ln(2)Ihalf-life ot,'radionuclide, 1/s;

mnatrix bulk den ;y-,.g Kd 2 equ i Iibr lum distj butjio'n copjf i c Ient,. cm /g.

.The processes modeled ýnIEq'.ý 2 are advection', molecular d~iffusio~n, -ehncal- dispersion', radionuclide decay, adqulbimsrto. For - . - . :: . . . ... ' .- •.-.,

1t1 this. report, the TRACR3D m 1ode is used in its two-dimensional configuration.

The simulations were compute*iný the z, the vertical direction, and x, the "horizontal direction. Becaus:.we are considering one-dimensional flow in

the downward vertical* irectl•h`,- u u , Therefore, D aLU, and ( Bea 0e72).: Tb• t ' .. I-and II give the values orf C, 0, 0 and

K for each of the'stratigeaph l units. a) d -In :the/ calcul ion., `, . ha "" "e

CU trecarried out for this study, two constant concentration radionuclid..4, .I.sources, each 0.25 m wide, were placed 30 m

apart. The zone sizes, or .•gr ds#pacings, in the horizontal direction were

increased from 0.25 toonemoved away from the sources. The finer grid spacings near the sour-ce-v. alowed the region near the source to be well

characterized. In the horizon •a direction, 52 zones were used to model the

total horizontal distance, J72',75i.m. This horizontal distance was chosen so that spreading of the radioiL•:iUd'e plume due to lateral dispersion would be

modeled and included in th4s•l~rsults. In the vertical direction, 130 zones were used to model the distacýt jfrom the line source to the water table,

265.4 m. This number of 'vertic'al zones was chosen so that the number of

subdivisions per stratigraphid Unit provided a vertical cell size of approx

imately 2.5 m. The actual min.Imum and maximum zone sizes in the vertical

direction were 2.416 m and 2.786 M, respectively. Test problems were run on

15

,-the CRAY XP? computer to .nvestigate the effects of ione z on the c. cracy-b•. .h,ýrf~e'esu1ts. The, zone.: 12eSIven above;.were selected becausee

~ftrther .. iemn of the znsre.3ul ted-,Jn litei ny. changes, In the gi,ýeaerscal arswers.- '-".:.

B. Scenarios ,- o , o . h

The transport,. of uranium and technetium from, the, repos itory to. the water table was modeled:,by using the geochemicai,'geophysic-al model discussed Ij-: -v Section IT! as a bass Ia. Uranium is, representative o:• those radionuclides

that have longer, halr-lives. and are more' relactive, with:-the porous media. Technetium is representative of' those radionuclides that are relatively more, soluble and. less reactive with the.porous media. In the, calculations,. two

-:,ra.dionuli fde line sources of1 constant concentration were used. To inyestigate ,the' effects of flow on transport,:two flow~scenari :

we.re Included: :a possible expected vertical velocity, 05."5s myr1 and a p-tential extremevelocity, 1.5 mm/yr. Because of the unavailability of.

data and an incomplete.understanding. f the technet sorption cproessi ,._a-zdtional cal culationsw w ere ma de -whic'h nelcd Te sortnInalth

Sstratigraphic units ie., Kd(Tc) 01. Table I II Ilists the. six transport, c cases. examined in this report; The'results from Case 3 are.compared with

iose.of.Case 5andthr tof Case 4 are compared with those of Case 6

to investigate the effects of sorption on transport. The results-,from Case .. are compared with those of Case 2, the results from Case 3are-compared

i with those of Case1.-, and the results from Case 5 are compared with those of : ! Case 6 to Investigate the effects of flow on transport. Case 6 represents

aýn extreme-case scenario for radionuclide transport from the repository to

. the water table. The extreme-case scenario has the higher flow rate of 4.5 C 'm.•/yr and zero sorption of T

V. RESULTS

A. Uranium Transport

Figures 2 and 3 show the results of simulating the Case I scenario "(Z-U transport; Uz: 0.5 mm/yr). Figure 2 presents the logarithmic con

centration profiles of aqueous-phase 2U. Figure 3 shows the logarithmic

• concentration profiles of solid-phase ...U (i.e., the concentration of 23'U

16

-7 T .... .

.sorbedonto thex tuff) Profiles at tmes of approximately 5,0- 0 '1"00000 and .15,000 yr *are, shown. The, leading edge. of the •"U. plume migra-ted ap- ..

,prox.1ma:tely ý20 m In `-00yr'In. the paraeraphs-A'tia. fol low1 theý lead'Ing

:e-e or the ple ned as the line with a .-t on 0f:- I, O': radionuclIde/c of li quI Io At: 10,o000:and 15,000yr, the. aqueou3m and,'

sobdphase U~t remained in the TSw2 unit-and had, not mI'rtdit h

Tuff of, Cal'c-3 Hills. . <. . *.*....- 7 jq'A-%. * Figures 4l and 5 show the results Of simulating the Case 2-scenari&'

(23$U transport; U.z 4.5 =/yr)..- Figure.-I presents the logarithmic-'concen...

tration. profiles of' aqueous- phase "'U. Fi1gu•r•es]'5 shows the logarithmici ' con. centration profiles of solid.-phase ,U•.P•rofil.es at times of approx- ..

Mately 5,000, 10,000, and 15,00 yr. are• shown., -in this case,ý Under the, con

ditonofan. extreme fl~ow..rater the. Uedn deo h " plume migrated Sapproximately 40m in 0,000 yr. At. 10,000 and 15,•000 yr, the, aq us- and.

"sor bed-phase 2 3 'U remainedý contained in -the TSw2, unit' and had notmigrated

into the .Tuff-of Calico. Hills. Figure 6& presentsa comparison of the. results'of the U transport'

calculations at 10,000 years- for both Tflowý scenarios- (Cases 1- and 2) . In

Case 2 where the flow. velocity -is set -at a. potential extreme value,. the

leading edgeof: the_plume has migrated .through the: reposltory .only. -20.m more than in Case 1. The increase in the flow by .a factor of 9 (almost an order

of magnitude) resulted.:,in only a factor of 2 increase in the distance the

leading edge of the "U Plume-migrated. The simulated distance the plume.

migrated was only slightly sensitive to the magnitude.of the average flow.

"In the case of. uranium transport, equilibrium sorption had a more signif- - -:

icant effect in controlling the long-term transport. In the TSw2 unit,

Kd("'U) was estimated to be 2.84 cm /g. This Implies that.for every 10: units of uranium in the aqueous phase, there will be approximately 60 units,

of uranium sorbed onto the tuff (see Eq. 1).

B. Technetium Transport

Figures 7 and 8 show the results of simulating the Case 3 scenario

(19Tc transport; uz: 0.5 mm/yr; sorption). Figure 7 presents the logarith

mic concentration profiles of aqueous-phase 9Tc. Figure 8 shows the

logarithmic concentration profiles of solid-phase 9Te. Profiles at times

of approximately 5,000, 10,000, and 15,000 yr are shown. The leading edge of

17

the 'Tc plume had migrateda pproximately 25 m at10,OOC yr. Underthe d.i.ion of an expected flow., rateof 0.5.-/yr, the aqueous- and. sor5.m/ha

pe lumes remained-in theT.Sw2ý un-it and did not-migrate Into the Tufof' calico 1:kls. The 'resUlt's for: ''T transport: closely-.resemble the transport results simulated using the low- _flow sc'enario6 (se.e Fig. 2)

Figures 9:and 10:show the, results of> simulatin the Case As n-o~<

"reT transport; us 2-4. 5 mm/yIr 3 sorpt Iion). Figure,9 peet hlgrt~< mic concentration, profiles of- aqueous-phase Tc. Figure 0 Wshow

logarithmic-concentration profiles of sol.id-phase .Tc.- Profils

of approximately 5,000, l10O,0, and 15,000 yr are shown. the leading edge, o

. the ".Tc plume had migratedý approximately.65 m at OO00 yr and remained

.ithin the .TSw2 unit t is interesting to note the behavior. of •"'Tc "

15,000 yr. At .15,000 yr, .the..1Tc-plume had migrated,"into, the Chn zu

apprcxlmately 125 m. below the source.. Below, the TSw2. u nit -,- the plume does••:-d

r.ot exhibit the extent of the lateral spreading observed& in the 7Sw2-uni

"7The values for distribution' coefficients Afor the units underlying. TSw2 . ra 0.0 in TSw3, 0.04 in CHnlv, and 0.0098. in CHnlz (see- Table i1). Sorpti6rj.ý7R> has .an negligible effect on transport in these-units.'- However, Fig. 10 does!

show a -orbed-phase concentration, of "To in, the CHnIV unit where K' O'z 04,,' /red Lith the low flow scenario where the results of- I, Td

cm~ /g. Compare here&u "2 U transport closely resembled each other, in the extreme flow. scenario

the 19Tc plume migrated 25 m farther than did the ) 'U plume. As the:fl

rate is increased, sorption becomes a less dominant mechanism in cont.'roi

transport.

Figure 11 presents a comparison of the results of the •T transo

calculations at 1O,000 yr for both flow scenarios when sorption is con

sidered (Cases 3 and 4). In Case 4 where the flow velocity is set at ' potentially extreme value, the leading edge of the plume has migrated -.

through the repository 40 m more than in Case 3. An increase in the flow-,

a factor of 9 (almost an order of magnitude) resulted in an increase in7.the

distance the leading edge of the Tc plume migrated by a factor of three

The simulated distance the plume migrated was moderately sensitive to the

magnitude of the average flow.

Figure 12 shows the results of simulating the Case 5 scenario (''Tc

transport; u 1 0.5 mm/yr; no sorption) Although Cases 5 and 6 may not

18

CNI .Di

represent actual system behavior.j or.,,c. :there is still uncertainty as

sociated with the. sorption behaI 2.-fin the TSw2 unit (Thomas 1986). 7ne,,result; from: tha lmul a tf-"hesults from Case 6&are .representative of thetrnn

ofonserv�aa ive, nonsorbing radionuclides and are -helpful--in exa•i• in t ete sorption on transport., Figure 12 presents -the IoiarIthMIc concentraltn€ I on iles Of aqueous'-phase .Trc.

Profiles at tiMt3sof, approximately . 5,QO10,O000, and 15,000 yr are shown. At 10-,000 yrLhe leading edge ot ^plume migrated out* of the Srepository unit. and •into the CH6v 'unit, 6pproximately 100 m below the source.. .

Figure 13 shows the results ot s "ulating the Case 6 scenario ("To trnpr;u:1. r/yn opin.zcs 6. represents an extreme...ase

transport scenario where the higherAflowrate, 4.5 r.n/yr, and zero sorption of .9Tc are combined in oe -sim jto - Figure 13 presents the logarithmic oncentration profiles of aqueous Phas.e2 Tc. Profiles at times of ap

proximately. 5000, and 10,000 yr are At 10,000 yr, the leading edge of thee Tplume migrated througt i't:re- unsaturated zone into the CFUn Ustatic water leve .totali ng a distance of approx

imately 300":m .

The only.case in-which. any caiculated amount of radionucl ide was transported to the water table was Case... 6. -In Case 6, approximately 1.6 x

0" gm (2.8 X16Ci) of •Tc/twb:o rees was estimated to have accumulated below the static water" tabl'e hi" the groundwater at 10,000 yr. For one line source, this equals 1.4 x IO j",,uree. The allowable -umulative release to the accessible environimentý"tfr-.10,000 yr after disposal is 10,000 Ci of Tc per 1000 W (me t r ica -tos'li6theavy metal) exposed to a burnup between 25,000 and 40,000 megawatt-dy. r (U.S. EPA 1985). If there are

70,000 KMTI in the Inventory (p. 6 -365k V.'S. EPA 1985) or 27,000 canisters each containing 2.59 HTHM in thee potential- repository, then the EPA allowable cumulative release would be 700',.- •-Ci of "Tc in 10,000 yr. Given the conditions of Case 6, assuming, one sicource represents a canister, and assuming all the 27,000 canisters in- " e repository were to be breached, the total curie load to the water table would be 3.8 x 10- Ci, less than 10" (or one hundred millionth) of the allowable EPA containment requirement.

Figure 14 presents a comparison ofthe results of the "Te transport calculations at 10,000 yr for both flow scenarios when sorption is neglected

(Cases 5 and 6). In Case ...ere, the flow velocity is set At a potent ial extreme value, the leadingedge of, the Pliume has migrated through-the repos Itory .200. M ma`re than in' Case MEA nraei hefo yaf~o' of 9 resultted in an Increase in' the dis e th e

-plume migrated by a facor a of.3ý In this comparison where sorption Is&z the simulated.distance. the plume migrated was moderately sensitive toth magnitude of the average -flow. Advect ion is the controlling trans~r'-~

" •- p ces'..

C. ffects of Sorpt ion

The. effects, of sorption. on the transport -of radlonucl ides can be ex amined by comparing Case 3 with Case 5 and Case 4w.th case 6 .Flr 15:';•

:shows the comparison of results of th a-9

apprximtel 10000 r fr c* fthetransport clulations for 'Tc t. S approximately 10,000, yr for :Cases.3: a~nd.5 The.s5e flow: scenario,0 rrm/yris used in the two cases,-t. btin Case 5 zero sorption:of, "T.T- is assumed.: In Case '3, the. Ta plume, re mains within t TSw2u unit. -- In Ca s- i

5, the leading edge'of. the. plume mirtd iote, C~n1v .unit, 80 ~ a fac-~

tor of. 4).farther than did- the plume in Case 3. ,,Also note the Increased

lateral spreading in. Case-5, which is due: to lateral dispersion.. .. .. Figure 16 shows the results of transport calculations for •T: at: -i

approximately 10,000 yr for Cases 4 and 6. The same.flow scenario, 4•5 =/yr, is used in the two cases but in Case 6 zero sorption of "TO is assumed. In Case 4, the To plume migrated 60 m and remained within the TSw2 unit. In Case 6, the leading edge of the plume reached the static water level, migrated into the CFUn unit, traveling 230 rn-(a factor of 4) farther than did the plume in Case 4.

In these expected- and extreme-case flow scenarios, the estimated 9transport was very sensitive to the degree of sorption. In both com- -: -. ; parisons, when sorption was zero the leading edge of the plume traveled a. factor of 4 farther. Sorption can be a significant retarding mechanism whenýthe flow is at either a relatively low or high level.

" AI

20

S... . of0 FloW J:•

D. -Ef recot: flw

The. effects of -flow~on the -transpor tý ofý radIonuci Ides can be summarIzed by compa.ri Case )with Ca 2. (it Case (Fi. :1.1. and

Case 5 with Case 6 (Fig. 14). These comparison3'ýwere described in detail In

the'previous sections~. In these three ýserts,,,he~ sorption was held constant Vin -each set and the flow was chang~ed from 0. 5m=/yr to .l.5 mm/yr. In Fig. "6,"" the• increase in the flow by a, fa:t ?ofi9resdlted In a factor of 2

'incra. ig of the tU plume migrated. The in a ein- Ie distance tt i, hed-leading

actual;diffrnei-dstnemgae was:2 m., In Fig. 11, thledn 19

edge: of' the iTo plume migrated 40m more (approximately a factor of 3) when

the higheri flow scenario and sorption were':onsideed. In Fig. 14" the,

l eading edge of the. 19T plume nigrate' 206sm moz-re. (a factorof 3) when the

-higher flow scenario was consideredan d sorption was neglected. As the

am".oun t- of sorption decreased from cmI-m parisoato comparison (Z U, Fig. 6;

"49Te with sorption', -Fig. 11; .T. zero sorptioni Fig. 14), the actual di f

"ference In distance the leading. edge of,6 t plume migrated increased. When

-.sorpion i a significant controlling "ransport process as in the case of

"'U transport,:: the transport is only slightly sensitive to an increase in

flow., Wen sorption "becomes less signiiicant.the transport becomes more

and more sensitive t_ýo an increase in fth•e"fl

VI. DISCUSSION.

A. Geochemical/Geophysical Model

As the Yucca Mountain site is more t"lly-characterized in future inves

tigations, the geochemical/geophysical m"6.delwill be updated and revised.

Several important areas need to be addres-sed.'4The stratigraphic model for

the transport calculations should in~clude,-, :the -5,to 10 degree eastward tilt

(Ortiz et al. 1984) of the tuffs and expliq"t.'Jaults. Geologic faults need

to be included so a more accurate model ial transport pathways can

be provided. In a recent report (Roulon' .eeiaL 1986) it was shown that

lateral flow caused by the tipped beds occurred. under both matrix and frac

ture flow conditions. The tipped beds.,.and.geologic faults were not included

at this time because of the unavailability,,o data concerning the natural

groundwater flow and fault parameters. .

21,

?E

•-.-:.S.:":-::.:>" '- :'>=, . " "."" -. . '-•3;•;• - : ..' .o ' -. -'A . - -;. . '•

- ' ' .- .. *•

A better underztAnd~irij o the natural groundwater flow in each-of the strat.Igrahiun.its.j&'. Work' In this area ialso in preliminary . s-ae' e., Figure, 17 shows a hypothesiz ed model of the ý'"r .egcime through ,.th-byd geologie units at Yucca Mountain. If Indeed a .portIon of theltowu •dterteid•y laterally around the Tufr or Cali o fHll-. and intoa much ultzone,. the. groundwater travel time,. to the accessible envd.onmenttu be decreased and radionuclide, releases cduld be increased&Y..

There 3i:s4 general hi'iitlability of data for most of the parameters that describe the logiamedia. Much of the available data is extremely locaizedaround the drlf-holes where the core samples were collected. Some-type. of data -analys Ii,,ch as krfging, should be done on :parameters

- " suc as:saturation, ro,_.y t~nd bulk density.. Values foi those parameters. could. then"be•er at'b. ploa..d....locations away from the drill holes, and estirates.of error-could. ba•"...,signed to the estimated parameter Values.

inBecause sorpt ionplapsuklan important role in the overall estimated . .rt mor - . t•eeaus" e- s,.n

C)tranpo mre soio dataýr•elated to the units directly underlying the repostor should be- col e Also, more information is needed on the . -. reitsitorytshoerrors associated tho.experimentally determined distribution coefficiehts.' "ore.zpeoHf:ft geochemistry should be investigated such as the ~ geochemical... .--:speci~es _,. re ia .'::tthermodynamic formation constants for aqueous

cc, and sorbed.species, a .... E.native: sorption models (e.g, isotherms and nonequilibrium sorption, mdeli)V'

C-'-"" -unless it can,.be.s_.: nhat data can be inferred with a high degree of confidence, less -unceirt-a Inj and more material and geochemical data should be gathered.'. In some cases'im'uh data exist within one layer, but none exist within the-next layer t'i,,`.v.emonstrated with the sorption data, several stratigraphic 'units sociated values of distribution coefficients.

B. Transport Calculations)tx%§

As the geochemicaigeophysical model is updated and revised, the baseline transport caleulatfons will also be updated. The best available set of transport, estin es .needed as a baseline from which to further investigate the effects3 and;.significance of the geochemical and geophysical processes controlling transport.

22

:" : iv particular, as the natural groundwater no•lbecome -better :/i8d!tood, more complex flow models need to be Incorporated into the S" " ans.-.. calculation. Lateral flow and0 possile fl6i. throth 'more permeA~~al~falt oneS should be. added. fflw 3s diverted .L~r~ o af

" e,,ra nsport. path t6 the accessible environren• tmay be shortened .and cumulative radionuclide. releases may, be -increased.- The transport should be -estimated at other locations within Yucca Xountain. -inti:nodations, the di`t

therlocah ditance to the water table, is much-less than in the. examples. presented h6ere and the material unIt properties mray.:d if fer (Ortiz et~al. 19841). ::.The effects of dispersion on the cumuiatve radionuclide release should "a-so be investigated, Dispersion is responsible, for the mixing and spread

-. :"'i"ng- of: Ythe. plume caused- by- micro6scopic velocity. variations within the pores and..I by large-scale heterogeneities"withIn tha geologic media. The'dsper_ CV-'4ii~y iin these examples waas set to a mi:i-mum. value:. However, as the

-' dispersivity increases, -:one.'might expect an increase in the cumulative release caused by a greater spreading of the radionuclides in the-direction

-Iln summary, theý calculations presented in this paper are preliminary in:

.~ ~ ~ ~ ~ ~ ~~~~r ,...::..) :,....-.., .cplya .... o.-'. _:.

" •.nature and serve principally as one example of how radlonuclide transport "".- be investigated" Before results of these calculations can be used in

;,-Judging the suitaility of the site for nuclear waste disposal, several facn need to be addressed. First, an accurate and referenced foundation for S any,-et of, transport calculations should be compiled. The availability and ,:,iel l•bility of the data should be assessed. Areas where more data are re-quired should be brought to the attention of other investigators. Finally, • he:results of many scenarios should be examined to isolate the effects of anY;'ne physical or geochemical process affecting transport.

- :: t CONCLUSIONS AND RECOMMENDATIONS _is report presents the results of a first set of calculations of " .lide transport from the disturbed zone to the water table. The

transport pathway is the unsaturated zone. The geochemical/geophysical mde ."used in these calculations is a preliminary one; much more information is .needed before the geochemical/geophysical model can be regarded as sufflciently complete so that it can be used for making dependable predictions.

23

The transport of two radionuclIdes is modeled: 231U and "Tc. Uranium.. is representative of radionuc¢idez with long-halt-lives and high values of. sorp~tion coefficients,_.and_ techneti.um -srepresentative of the most soluble' and fastest moving radlonuel ides..TWO T-fl0 Veloi.•ties, through the tuft are- . used for each-radionucllde and f'or each s3orption. condition: -an expected. average velocity of.O.5 mrin'/yr and an ext:reme :.average velocity of:.5 mm/yr. K " Thus, six transport scenarIos .are modeled.

A,. -Conclusions

The following major concn'lausions were made from these simulations: (1) For uranium, the estimated trran-sport is only moderately sensitivto the magnitude of 'the flow. The insensitivity -is caused by the. sorptive properties or. the••tuffs oft., Yucca Mounlain. (2) 'For technetium, the. sorpti'on distributiOn coefficient is only: slightly less than that for'uranium in the repository unit, and the:. results. for technetium transport closely resemble the results for uranium transport. (3) Flow dominates the transport: when sorption is .negligible, (4) The uranium, based on the preliminary geochemical/geophysica . model, is not transported beyond the stratigraphic unit In whch. 'the: repository Is 1-oated in both the expected and extreme flow scenarios. (5) The technetium, based on the preliminary geochemical/geophysil-. model and preliminary sorption data, is not transported beyond the stratigraphic unit in which the repository is located in both the expected and extreme flow scenarios. (6) If technetium sorption is negligible, a measurable amount of it moves to the wa,'er table under conditions of high flow rate. Nevertheless, the amount of technetium transported to the water table, based on the preliminary geochemical/geophysieal model, is still within the EPA-defined limits for a nuclear waste repository.

B. Recommendations

The following recommendations for future studtes and data aquisition are made to better determine the significance and Importance of processes affecting transport, to more fully characterize the site, and to assess expected postelosure performance of the potential repository.

24

-.--o Xany'mo re

sat

transport~.

(6) Addil

the flowI

out.

led.,before, the g'-ochemical/geophysical m Iodejl.

;arded as -satisfactory and a suitable b as e:

7-oI~~nof these data, should be a highi

ýsport 13 so sensitlveto" sorption, mnoreK

rradionuclide sorption, on, the varilous tufrs

ly the 3tratigraphic~ uinI it directly underly- .

eded on. the uncertainties associated with

sorption coefficients...

yss (e.g., kriging) should be done. on the

orptIc-.,c , "e:ci, lent data.'.

Ion and matrix and/or .fract-ure flow on gated.

imulations coupled, to more'complex-rodelso o

aphy. within Yucca Moduntain shoul~dý be .carried

25

9.

C

Sea r, J., 1972, Dynamics ofl Flidsinj Porous Media1 American Elsevier, New * York, 1972, pp. 579-q17.1e ,

S Broxton, D. E., 1985, "Chemical V ar ia biriVy of' Zeolites at a Potential Nuclear Waste Repository,. Yucca: Mona, Nead, trsne atth * I ter ati nal Sym osi m o Maagement of ýHazardous: Chemical Waste Sites

Winston-Salem, C

-.Broxton, D.* E. , 1986, Private Commu n ica t Ion. Los 'Alamos National Laboratory, Lo3 Alamos, NH.

*Campbell, K,., 1986, "Kriging :for Interpolation of Sparse and Irregularly Distributed Geologic, Dataj" NNWSI Milestone 376, Los Alamnos National Laboratory, Los Alamos, NM.

Gelhar., L'. W., and C. L.: Axness, 1983P "Three-D imens Ional1 StochAstic Analysisi of' 14crodispersion iAqfes"Watter Reeach191) 161.-183.7

Greenwade, L. E., and G. A. Cederberg,1 1987,. "Preliminary Geocheia/epyia Moe fYca Mountain," 1.NWSI Milestone R3143, Los Alamnos National Laboratory, Los Alamos, NM.

-'-:Kerrisk,' J. F., 1985, "An Assessment of the Important, Radionuclides 'ini Nuclear Waste," Los Alamos National Laboratory report LA-104114-MS, Los ýAlamos, NM.

Matheron, G., 1963,_"Principles of Geostatistics," Econ. Geol. 58L 121461266.,

-~Matheron, G., 1971, The Theory of Regionalized Variables and Its A~pplications, Ecole des Mines, Fontainebleau, France.

S Matheron, G., and G. de Marsily, 1980, "Is Trar sport In Porous Media Awy Diffusive? A Counterexample," Water Resources Research 65)90-917., -Montazer, P., and W. E. Wilson, 19814, "Conceptual Hydrological Model of Flow in the Unsaturated Zone, Yucca Mountain, Nevada," Water-Resources

-~ - investigations Report 814-43145, U.S. Geological Survey, Lakewood, CO.

Ortiz, T. S., R. L. Williams, F. B. Nimick, B. C. Whittet, and D. L. South, I p 19814, "A Three-Dimensional Model of Reference Thermal/Mechanical and Hydrological Stratigraphy at Yucca Mountain, Southern Nevada," Sandia National Laboratories report SAND814-1076, Albuquerque, NM.

Oersby, V. M. , 1986, Private Communication, Livermore National Laboratory, Livermore, CA.

oversby, V. M., and C. N. Wilson, 1986, "Derivation of a Waste Package Source Term for NNWSI from the Results of Laboratory Experiments," Proceedings of the Symposium on the Scientific Basis of Nuclear Waste

26

Management IX, Stockholm, Sweden, Materials Research Society'; 'Pittsburgh, PA, pp. 337-346 . .:ý

Pickens, J. F., and G. E., Grisaki,: 1981, Scale-dependent,'Dispersion in a Stratified Cranular Aquifer," Water Resources Research 17(_4). , . 1191-121t.

Feid, R.. C. , .1. H., Frausnitzi and T'.. K. Sherwood, 19,77,' Properties or Cases, and Liquids, 3rd ed., Mc•ra4.w. Hill- Newý York, pp. 576-577.

Roulon, J.-G,, S. Bodvarsson,h and .T. Montazer, 1986, "Preliminary, Numerical Simulations-of Groundwater- Flow in the Unsaturated Zone.-,Yucca Mountain, Nevada," Lawrence Berkeley" Laboratory report LBL-20553,: -Berkeley, CA..

RIndberg, R. 5., 1986, Private Communication, Los Alamos Naktional Laboratory, Los Alahos,• NM.*_.

Satterfield, C.. N., 1970.. Mass Transfer in Heterogeneous Catalysis,-M.i.T. 'Press, Cambridge,, Massachusetts, po. 33-4=1.

Tien, P., M. D. Siegel;. C. D. Updegraff, K. K. Wahl, R. V1. Guzowski, 1985,• :: "Repository Site. Data".Report• for Unsatkrated Tuff, Yucca Mountain, Nevada," Sandia Natiornal Laboratories. report Sand 8Z4-26 6 8i . Albuquerque,

Thomas, K. W., 1986, "A..Summary Report on Sorption Measurements Performed with Yucca Mountain Tuff Samples and Water from Well J- 13, .NNWSI Milestone M313, Los Alamos National Laboratory, Los..Alamos,- NM.

Travis B. J.,, 1984, .TRACR3D: A Model of Flow and Transport in Porous/Fractured Media," Los Alnmos National Laboratory report LA-9667MS, Los Alamos,-N.

TJFF Data Base Product ICALOOI0, 1986a, Sandia National Laboratories, Albuquerque, 11M.

TUFF Data Base Product 15, 1986b, Sandia National Laboratories, Albuquerque, NM.

TUFF Data Base Product #4, 1986c, Sandia National Laboratories, Albuquerque, NM.

U. S. Environmental Protection Agency, 1985, "40 CFR Part 191:.Environmental Standards for the -Management and Disposal of Spent Nuclear Fuel, HighLevel and Transuranic Radioactive Wastes," AH-FRL2870-3, Washington, DC.

Wilson, W. E., 1985, "Unsaturated-Zone Flux at Yucca Mountain, Nevada," letter from W. E. Wilson (U.S. Geologic Survey, Denver, CO) to D. L. Vieth (WMPO, Las Vegas, Nevada), December 24, 1985.

27

snflflflPflfl�TSr

%C

C',

ell

28

-M era- jrpe.' 3't St at

I t ~ t - S ra igraýh ic' Uni ts'

-Sa turattoI : 16a p orosjty b-Bul. k Dens .ityb unit' / gcr)s

TCw 01' 861'41: NA 2.2 497,, P~a 0'.241- NA fl421 TSw 1 01.81527 - 0.11681 2d1672'TSw2 :0.885'68 0 .11681 2.2629

Tsw3 '0:.75.876 0:11681 ,2.2953 CHnlv. 0. 76,1 0.35105 1-.5339 Cflnlz 08896,0: C.30636 1-.5910CHn2. M.11 38 0.1306-16 '1.7750 CHn3' 0. 940O3, 0.O306-25 t 1-.800 P~w -p.84744 .0.252741 1 * 8782 CFUrv 0.92025 0.32393 1.6391

a. .TUFF -986b b. TUFF 1986c

U,co N

.1

U

*�0

.0

'I

4)

0 4.) &,

- Cd

4)

if 0 S

-4 %- -4 .0

4.�. I.. 5s� -4 6.

41 U'.

:2 C.-. 0

C~ -. ..c

0

I~ .0

I1 3, , .

71 t-

- - vs>.

- -t

9'

-'I',

ýi , 5ý G (

00e.0

a,.00a

~.F' F;CF Fj

.0 t'¾. 0n LA "

*F~FaF0FUy U0 * ~ r "0-:27$c. >h

.~~FFF) 'FfjF 'Fgsj 4FF FQ'$

F 'FF F', F., ~ 'h'' ''' F'FFFn

F, F.

0 en

iUL

Etv

T~w

PTn.

TSwl

TSw2 - Sa

Ttw3. .

'Chnlv

CHnlz

P Nw

CFUn

0in Ur

257.E62VI~

13 1.7

.77,11

Figure 1. Schematic representation of' Yucca Mountain stratigraphy (not to scale).

I,

it-'

'-I I-'

C(2

SWL

238 U (glim)

5006.9 yrs 0.5 mmlyr

tFJ

- e20 4 so X-direcfion (rmeterI)

238 glmO10005.5 yrs 0.5 mm/yr %" J(g/mI)

j 8_ . . . . . . ,,I. . . . . .._ _ . . . . . 0 20 40 B • 0 20 .40 6o

x-drctio-X-dict (meters)

Figure 2. Liquid-phase concentrations durInfl.gtranlsport or" through un

I saturated Yucca Mountain turr at, normal•f flow rate.. Adjacent . ine. differ

in concentration by a fac utor or 0 0. O r/yr. 5000, 10000,"

and 15000 yr. "..(Case 1)

15004.2 yS 0.5 mm/yr Sorption

C)

Iv,

S,'. -•.

2 3 8 U (gig)

5006.9 yrs 0.5 mm/yr Sorption

0 20 40 40 X-direclion (mefers)

238 U (gig)

A -,

_E

4=,,

N

0

a0

10005.5 yrs 15004.2 yrs 0.5 mmiyr 238 0.5 rnmlyt

U (g/g) Sorption Sorption - _____________

. - IiI -= A-

U,

4, 4,

0

U 4,

N

•0

0 20 40 60 X-direclion. (meolrs)

ao 0 " 20 40 , ,

X-dirctlon (m.,!='=)

Figure 3. Sorbed concentrations during transport of "'U lthrough unsaturated Yucca Mountain tuff at normal flow rate. Adjacent lines differ in concentration by a factor of 10. u = 0;5 mm/yr. t 5000, 10000, and 15000 yr. (Case 1) Z

�A½�tu

*2'J'(. �-. . -�

fri.

S

N

0

C,

Iv,

so~

N

0 ý-

. .,..

8 8

238 U (g/mO

5006.9 yrs 4.5 mm/yr

X-dfirecton (meters)

238 U (G/mD

so 0 S1 Q so

X-dlrdcton (mHIrr3)

2 3 8 u (/mD

I-. �

so 0d 30 .40 so

Figure 4. Liquid-phase concentrations during-transport of Z"IU through unsaturated Yucca Mountain tuff at high flow rate. Adjacent lines differ in concentration by a factor of 10. u = 4.5 mm/yr. t 5000, 10000, and 15000 yr. (Case 2) z

-S

S 0

.;..

* N

0

15004.2 yrs 4.5 mmn/yr:. .Sorption,

II i C!)

_- _ . . . -

•./ii

!

2 3 6U (g/g)

5006.9 yrs 4.5 mm/yr Sorptin

0 40 6 s X-diroclfon (motorl)

238U (g/g)

I

*0

U

N

a

10005.5 yrs 4.5 mm/yr .Sorption

Sorpt�n

A-

23 (g8 g U ....

8 'II

X-dlrecflon (mof-rs)

15004.2 yrs,

4.5 mm/yrS. .. 5 - . .

0 I0 . 40 O O0 -

Figure 5. Sorbed concentrations during transport of 'U through un-saturated Yucca Mountain tuff at high flow rate. Adjacent lines differ In concentration by a factor of 10. u = 4.5 rm/yr. t 5000, 10000, and 15000 yr. (Case 2) t 5 2

U

LA

0

T'R--

Tra

C

S. . . . . . J i l l

10005.5 yrs 0.5 mm/yr Sorpton

23U (g/mD

S 0 40 so

X-dircfion (maiers)

10005.5 yrs 4.5 mm/yr Sorpton

238U (g/mD

DII

C

U

C

0 20 40 1W. X-dir ctk., (me-rs)

Figure 6. Comparison of liquid-phase concentrations of..'U after transportfor 10000 yr at differentef1ow rates (Cases 1 and 2). Adjacent ' lines differ in concentration by a tactor of 10.

a' w•-u

N

U)

C

cx�

S-- w w I - .I T

.i./•:,•';"/i.

I

0 O O1

|C?'

* 7 � .. , 7 . /77 . 7 . 7-77-..7'.' - 77' '7 .77

* ,.7..'.,.7.. rn � A. �'' A 7777377 77 7777,7)9,,

.7 .7 *.�- 7777 77 , 7

7 . . '7 7 . 7 .

99 To (gig)

R.____

4998.0 yrs 0.5 mm/yr Sorption

0 20 40 6o X-direction (meoers)

99

a * L.

S S 4;;

e� ,7*'4J7W'

- I N

.9994.7, yrs 0.5 mm/yr SorptionTo (g/g)

ao 0 20 40 60 X-direction (meters)

99 To (gig)

*777,7**

..7 7

7x7.,7

7,

7.-'

* 77

* N

0

w

A' 77' 77777\�777'j(..7,7�7.7

7 �7974 7 77 7� �7.7777 �4"'��4'4V.v

1499t.3 yrs 0.5 mm/yr Sorption

":.4

¶�9

'7 .'r'''' � 7 77777. �77 77 7 7

'U 7 7ft(737� � 773777

'7 7% �7477 '77.A ,74ft' 7'.f '$ t77L,;i$'7

.

77777 777 - 7','7

.',',7.r77 477 7

.74

774Y.

77r77.A

7737 77774.779.477777 '."-�'r.

so8 . 20 40 6o X-directio (m.iers)

7.. j, • .. .-. ,C , •q• ' •t . -'a,,- ,:• ,••

cc'

13 ...

• . .. . f . .22 i.'4 d.

Figure 7. Liquid-phase concentrations during transport of 'tTc with sorption through unsaturated Yucca Mountain tuff at normal flow rate. Adjacent lines differ in concentration by a factor of 10. u = 0.5 mm/yr. t = 5000, 10000, and 15000 yr. (Case 3) Z

8

9-, �7� IS S

N

a

ta -J

C.

. - . . -

0t

F

4998.0 yrS 0.5mm/yr SorptonTc (girmD

0 30 40 sO

X-direction (motors)

9 9 Tc (gl/m

.9994.7 yrs 0.5 mm/yr Sorption

20 40 go

X-directlion (meters)

9 9 Tc (g/mgln

149913 yrs 0.5 :mmyr

X-dlrectlon (motoers)

Figure 8. Sorbed concentrations during transport of through unsaturated Yucca Mountain tuff at normal lines differ in concentration by a factor of 10. t = 5000, 10000, and 15000 yr. (Case 3)

"IITc with sorption flow rate. Adjacent u = 0.5 rum/yr.

99

IO

'0

N

. 0

C7 r'•

;..�j,* ,�....

. . - =

' " " ... . .,• 7,S4

,A A:: . ' . :-, . . . . ' . /••

4998.0 yrs ,4.5 mm/yr .Sorptin

2D 4( etr X-directlon (mreiers)

9993.3• yrs 4.5 mm/yr' Sorptkin,'

9 3Tc (g/ml)

'4 L.

N

0

sI

X'dlrecilon (meters)

9 9 gi -Tc-(g/rnl)8 -

'4

. k . • . •

,*, .N .

N

Do a 2 0 4D 0 X-directlon (mnefrs)

Figure 9. Liquid-phase concentrations during:.transport tion through.unsaturatedYucca Mountain tuff. at high lines differ in concentration, by a" factor of' 10. u t; 5000, 10000, and 15000yry (Case'.4) Z

of 19Tc wIth sorprlow rate. Adjacent W .5 mam/yr.

-, S -�

9 9 Tc (glnl)

%P

14989.8 yrs A.5 min/yr Sorption

�0

N

0

C)

CO

0%

•L_ _ . - . •

.• _ w ,,

A • A I A • A I A A A J A • . . . . . . . . . . . . I

P. .Q .

u _) I. • ))

8

, UU

4998.0 yrs 4.5 mm/yr

20 40 X-dkectfion (meters)

9 9 Tc (g/g)

4' 4,

U S

N

0

9993.3 yrs 4.5 rmm/yr Sorption.

8s 0 20 40 60 X-direction (meters)

99 Tc (gig)

L S 4)

N

0

s o 0

14989.9 yrs 4.5 mm/yr Sorption

Sorptkxi

X-0 d 0dl : ( 0 m,..rs) X7 directlon (trr ters) ' •:::

Figure 10. Sorbed concentrations during transport through unsaturated Yucca Mountain tuff at high lines differ in concentration by a factor of 10. t = 5000, 10000, and 15000 yr. (Case 4)

of 9Tc with sorption flow rate. Adjacent = 4.5 mm/yr.

9 9 Tc (g/g)

A l -

S S

.2

0

0

tr0 0

CD

fr%

• - . . . . •'=' 'A . . .

.... ..... . ...... • . . .• -.. , ..

I

N

9994.7 yrs 0.5 mm/yr Sorptan

9993.3ý yrs 9 9 - g 4.5mm/yr

E$orptk'.

0 0 40 so s

X- .recft (melors) X-arecfion (mfrss)

Figure 11. Comparison of liquid-phase concentrations of "9Tc after transport with sorption for 10000 yr at different flow rates (Cases 3 and 4). Adjacent lines differ in concentration by a factor of 10.

9 9Tc (g/mf

i

ý L

4996.7 yrs 0.5 mm/yr No Sorption

9 9 Tc (glmt9 9Tc (g/ln)

9993.2 yrs 0.5 mr/yr No Sorption

0 20 40 60 80 0 40 60

X-drecfion (meers) X-direcfloni (meters)

9 9 Tc (g/mln.

9-.. a,, I

U

N

a

0 "

Figure 12. Liquid-phase concentrations during transport of 92Tc without sorption through unsaturated Yucca Mountain tuff at normal flow rate. "Adjacent lines differ in concentration, by a factor of 10. u , 0 •5 umn/yr. 6 000, 10000, ,and' 15000 y rr (case 5)

14990.0 yrs 0.5 rnniyr No Sorpqon

.2

;�'--'e �

I ,

CD

a-

20- 40 d n(

I. ,

m - , l

* - � *,�'

� �

99

4996.7 yrs 4.5 mrnm/yr No SorptionTc (g/mD

9 9Tc(ghnl)9993.2 yrs 4.5 mm/yr No Sofption

*1 L 0 C

U C

tO

N 0

0 20 40 W X-direction (moters)

C. L S S

N

0

ao

Figure 13. Liquid-phase concentrations during sorption through unsaturated Yucca Mountain Adjacent lines differ in concentration by a mm/yr. t = 5000 and 10000 yr. (Case 6)

a di 40 so X-di',,ctin (Mew#,N=)

transport of 9iTc without tuff at high flow rate. factor of 10. u = 4.5 z

* . .

I

I

9993.2 yrs 0.5 mm/yr No Sorpton

9 9 Tc (O/mft9 9Tc (g/m):

9993.2 yrs 4.5 mm/yr

No Soqption

0 30 40 00 so 0 40 so X-direction (motors) X-dir.ction (mnoors)

Figure 14. Comparison of liquid-phase concentrations of "9Tc after transport without sorption for 10000 yr at different flow rates (Cases 5 and 6). AdJacent lines differ In concentration by a factor of 10.

k

9994.7 yrs 9993.2 yrm .0.5 mm/yr0.5 99... . . 9 9 Tc (g/ml) Sorption Tc (g/mO) 'No myr

. • " : 'r '- ".'" "" .

N N . ,,.. .. . " .. ..

20 80 0 30 O @0

X-dkvcjim (m..rs) X--*Ofon (m.$.rs)

Figure 15. Effect of sorpt ion on liquid-phase concentration of Tc after' transport through.Yucca Mountain tufrfat normal".flow rate for 10000 ' yr.,, Adjacent lnes differ. In concentraton. by a factor of 10. u 0,5 .r•• .. r.m., 5) nd

,""

'-- :- -'-:...;•..U••S,-.,'.•'-.-".•-,'.,"•o"•"•'•".',;,••: •,w •.' -- '•-: -"• ',. • •;." • .. "• "." •' •".. ... " ;•" '"V:'!:. : •"; ';.':,• :.•• •:,'•;-• "

So�n

9

20 40 6O SO X-dfrechon (meters)

9 9 Tc (g/mI)

9, '

0' L S S

N

a

20 40 so X-dir-cI•on (me.wrs)

Figure 16. Effect of sorption on liquid-phase concentration of * after. transport through Yucca Mountain tuff at normal flow rate for 1.0000 yr: Adjacent lines differ in concentration by a factor of 10. u z 0.5 mm/yr. (Cases 4 and 6) z

9 9 Tc (g/rrd)

a.'

N

9993.2 yrs 4.5 m1mn/yr No Sovpion

If)

If)

If,

01

0

v v q• • -- NI

9993.3 yrs 4.5 mm/yr Sorpio

f

17'0;ý

e

.NOT-TO SCALE

ALLUVUM

TNA CANYON WELDED UNIT PAINTBRUSH NONWELDED UN TOPOPAH SPRING-WELDED UN CALICO HILLS NONWELDED UI

EXPLANATION

CF CRATER FLAT:UNn

IT ~ ~ DIRECT$ ' ONF o LtOl r -• • DIRECTION OF. VAP

IPERCHED WATER,

'II I

Figure 17. Hypothesized model of the flow regime through the 'hd-,geo units at Yucca Mountain (Motazer and Wilson 1984, Fig. 141; reprintýedwith permission),

47

'ICF'

I',

A

TC

P

TS

CH

w IlENT

"INC,