PNL..7474

UC-814

• '_'e

II II .... IIIIII I

Three-Dimensional Modeling ofUnsaturated Flow in the Vicinityof Proposed Exploratory ShaftFacilities at Yucca Mountain, Nevada

M. L. Rockhold

B. SagarM. P. Connelly

II I IIIIrl I I II

April 1992

Prepared for the U.S. Department of Energyunder Contract DE-AC06-76RLO 1830

Pacific Northwest LaboratoryOperated for the U.S. Department of Energy

. by Battelle Memorial Institute

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of tile ,United States Government. Neitiler the United States Government nor any agencythereof, nor Battelle Memorial Institute, nor any of their employees, makes anywarrantyt expressed or implied, or assumes any legal liability or responsibility forthe accuracy, completeness, or usefulness of any information, apparatus, product ,or process disclosed, or represents that itsuse would not infringe privately ownedrights. Reference herein to any specific commercial product, process, or service bytrade name, trademark, manufacturer, or otherwise does not necessarily constituteor imply its endorsement, recommendation, or favoring by tile United StatesGovernment or any agency thereof, or Battelle Memorial Institute. The views andopinions of authors expressed herein do not necessarily state or reflect those of the

United States Government or any agency thereof.

PACIFIC NORTHWEST LABORATORY

opera ted byBATTELLE MEMORIAL INSTITUTE

for the

UNITED STATES DEPARTMENT OF ENERGY

under Contra ct DE-A C06-76RL 0 18.30

Printed iri the United Statesof America

Available to DOEand DOE contractors from the

Office of Scientificand TechnicalInformation, P.O. Box62, Oak Ridge,TN 37831;pricesavailable from (615) 576-8401. FTS626.8401.

Available to the public from the National Technical Information Service,U.S. Department of Commerce,5285 Port RoyalRd.,Springfield, VA 22161.

II ..... 1 'IIIH II ...... li _........ Hre" lit'"'' _l' _'..... llll'_l" Ulllllll_

PNL--7474

DE92 012319

THREE-DIMENSIONALMODELINGOF UNSATURATEDFLOWIN THEVICINITY OF PROPOSEDEXPLORATORYSHAFTFACILITIES AT YUCCAMOUNTAIN,NEVADA

M. L. RoGkholdB SagartaJ (b)M. P. Connelly

Apri I 1992

Preparedforthe U.S. Departmentof Energyunder ContractDE-ACO6-76RLO1830

" Pacific Northwest LaboratoryRichland, Washington 99352

(a) Southwest Research InstituteSan Antonio, Texas 78228

(b) Westinghouse Hanford CompanyRichland, Washington 99352

i_,;_,'_.,

Dlol Fd,:,L.Il t .....b_0t:::ll.ll.t!_I..L)(1)(..;LIML_fq]ltli;LJt'JLIMIILID

1!I .,,,,,,.,,,,,,,,.,,........,..,..................,,,,,......,,,,,,...................,,,,,........,...............,,,,,,_,,,,,,,,,,,........,,...............,,,,,,,......,,,,,......,,_,,,,,,,,,,,,,,,,,,,,

ABSTRACT

This report describes the results of a study to investigate the influ-

ence of proposed exploratory shafts on the moisture distribution within unsat-

urated, fractured rock at Yucca Mountain, Nevada. The long-term effects of

exploratory shafts at Yucca Mountain are important in the estimation of poten-

' tial waste migration and fate, while short-term effects may be important in

the planning and interpretation of tests performed at the site.

" Numerical simulations, using a three-dimensional model, suggest that

most episodic pulses of water from natural precipitation should not reach the

proposed repository directly through fractures or through modified permeabil-

ity zones around exploratory shafts, due to capillary imbibition into the rock

matrix. The results of these simulations also suggest that the ambient mois-

ture content and distribution in the vicinity of the shafts should not change

significantly within a reasonable time frame for subsurface characterization

work (20 years).

Relatively few data exist on the hydrologic properties of unsaturated,

fractured rock at Yucca Mountain. The simulation results reported in this

document are specific to the discretization schemes, boundary conditions, and

hydrologic properties that were used. lherefore, these results simply repre-

sent approximations of the potential effects of exploratory shafts on unsatu-

rated water flow at the site. Improved confidence and quantification of the

uncertainty in this type of modeling study wil be attained as more data

become available.

iii

Im ,r, .,ll,,,plplltrllu,,,,llll_ , r_l,iTii I rl, lilppl,, r_ lll' fir' , ,,llr,Ul_ljl lm "' "Iii' II ,r ' 111r q, iiHll qIl. lV q_qll,,q' ql!'l'rr_' " II I1 II rl, Iii ,, ,,," , I11_' ' plplll' i , ' ,,,HI11rll ,r_l_I_ Hl II1' "

,kNlihl_lUl_lllllldllJlilllmlhHLlJ,llMH_mlilhll_lim,llhml_,u_|ll,lllliMkadlllmllul,bldl_ldl_lll,L,, ,bdH, ,IlL .... ,1, ,,,

EXECUTIVESUMMARY

This report was prepared by Pacific Northwest Laboratory for the Office

of Civilian Radioactive Waste Managementof the U.S. Department of Energy.

This study was undertaken to demonstrate the capability to estimate the mois-

ture distribution within unsaturated fractured media in the vicinity of a

. hypothetical exploratory shaft facility (ESF) at the potential high-level

nuclear waste repository site at Yucca Mountain, Nevada.

+ The ESF is an important part of the subsurface-based site characteriza-

tion activities that are outlined in the Yucca Mountain Site Characterization

Plan. During the construction and operation of the ESF, methods will be

selected to minimize rock disturbances. However, some alterations in the

properties of rocks close to the excavations are to be expected. The objec-

tive of this study is to investigate the influence of these altered rock zones

on the distribution of moisture in the vicinity of the ESF.

This study is limited to investigation of extremely long-term (steady-

state) and short-term (on the order of 20 years) impacts of the ESF on the

moisture distribution in the unsaturated zone. The long-term effects are

important in the estimation of waste migr,_tion and fate, while short-term

effects may be important in the planning and interpretation of tests performed

in the ESF.

The PORFLO-3computer code was used for simulation of moisture flow

through the geologic units adjacent to the ESF. Rather than represent frac-

tures as discrete elements, an equivalent continuum was stipulated, in which

the fractured units were assigned equivalent or composite hydrologic proper-

ties. Explicit treatment of fractures is ,_ot feasible because of the

extremely large number of fractures contained in the site-scale problem and

, the difficulties in characterizing and modeling the fracture geometries.

A three-dimensional geometry was used for the simulations. The explor-

atory shafts and drifts and the altered property rock zones [called the mod-

ified permeability zones (MPZs) in the context of hydrology] associated with

them are represented as one-dimensional ine elements that are embedded in the

n_,.11,,[n ,rl _P.... +._pl,,+.... r,Hlpl'" r_', rllr_l, lip ,+rn,+n 11n,, 1,1, lr, ., ,q, ,, 'l;pnl + ,r, III 'ht" pl'"r'1,1' r, ..... Pl .... _,.r 'Pq1" H'"

general three-dimensional elements. The Ghost Dance fault, one of the

prominent structural and hydrologic features, is included in the calculation

domain. This fault is represented using two-dimensional planar elements. The

ability to incorporate one- and two-dimensional features within general three-

dimensional elem,nts is a unique feature in the PORFLO-3formulation.

The results of these simulations suggest that the MPZsaround thee

exploratory shafts may act as conduits for preferential flow of water in the

welded tuff units when locally saturated conditions and fracture flow occur.

However, given the moisture and flux conditions in these simulations, fracture

flow with an equivalent continuum approximation is limited to the Tiva Canyon

welded tuff unit. Also, with the initial moisture and flux conditions spec-

ified, the Topopah Springs welded tuff unit is sufficiently thick that most

episodic pulses of water from natural precipitation should not reach the

potential repository through fractures duc to capillary imbibition into the

rock matrix.

A potential for significant lateral flow above the interface between the

Paintbrush nonwelded unit and the Topopah Springs welded unit is shown by the

simulation results. Where laterally flowing water intersects preferential

flow paths or structural features, such as fault zones, locally saturated or

perched water-table conditions could develop, which could lead to fracture

flow in the Topopah Springs welded unit (i.e., the repository horizon). The

Ghost Dance fault was embeddedwithin the model domain for these simulations,

and was represented by using the hydraulic properties of a much more permeable

material similar to gravel. These hydraulic properties are estimates because

of the lack of site-specific data. The Simulations suggest that there will be

no appreciable change in the moisture-content distribution below the Tiva Can-

yon welded unit within 20 years after the construction of the ESF. Therefore,

even though excavations associated with the ESF will alter the fracture den- o

sities and permeabilities of the tuff units, the ambient moisture content and

distribution in the vicinity of the shafts should not change significantly

within a reasonable time frame for subsurface characterization work.

vi

_l_li_!', ................ 'I' ,r' ........ I ...... ,I ll_llI'............ ,I ' '!l....... _ " i,.... r','..........lp,"'"...... ' ..... II............fir_=,'_r',......_',...... _l,,_lllll_'"",'qrlf"_'I'_'_l,,.......iit,,,,,,_....

A relatively coarse grid was used for the model simulations, with the

shafts and fault embedded as one- and two-dimensional features within three-

dimensional computational cells. Therefore, the results from these model

simulations are not as accurate in proximity to tile shafts and fault as they

would be if these features were explicitly discretized in three dimensions.

In addition, relatively few data exist on the hydrologic properties of the

' vadose zone at Yucca Mountain. Considering the comparatively large scale of

this model, and theuncertainty of the hydrologic properties, the results

presented should be reasonable. However, these results are specific to the

discretization schemes, boundary conditions, and material properties used for

these simulations, and simply represent approximations of potential effects of

the exploratory shafts on unsaturated water flow at the Yucca Mountain site.

Postclosure thermal effects and vapor-phase transport of moisture caused

by heat generated from the potential repository were not considered .in this

study. However, these factors may have a significant effect on repository

performance. Therefor'e, postclosure modeling studies should consider noniso-

thermal flow and vapor-phase transport of moisture. The MPZs and shaft back-

fill material may become more important for future, smaller scale simulations

when coupled fluid flow, heat transfer, and mass transport are considered.

Future simulations with intermediate- and large-scale multidimensional

models of the potential repository should include site topography and surface

expressions of major fractures and faults to determine potential areas for

preferential local recharge. Then, spatially nonuniform, time-varying

recharge fluxes can be applied as surface boundary conditions to determine the

importance of time and spatial variation of surface recharge to the flow

fields in the natural system at depth.

vii

CONTENTS

ABSTRACT............................. ii i

EXECUTIVESUMMARY ....................... v

1.0 INTRODUCTION ....................... 1.1,,

' 2.0 MODELDESCRIPTION . . ........ 2.1q

2.1 GEOMETRYOF CALCULATIONDOMAIN ............ 2 I

2.2 HYDROLOGICPROPERTIES ................ 2 3

2.3 INITIAL CONDITIONS ................. 2 12

2.4 BOUNDARYCONDITIONS ......... ....... 2 13

3.0 COMPUTERCODEDESCRIPTION . . ................ . 3 I

4.0 RESULTSANDDISCUSSION..................... 4 I

4. I STEADY-STATESIMULATIONS ................. 4 i

4.2 TRANSIENTSIMULATIONS................... 4 12

5.0 CONCLUSIONSANDRECOMMENDATIONS............... 5 I

6.0 REFERENCES................... ....... 6 I

ix

FIGURES

1,1 LocationMap of the YuccaMountain Site, Nevada .......... 1.2

1.2 ConceptualExploratoryShaft Facility. ............. 1.3I

2.1 _,YuccaMountain PORFLO-3ComputationalGrid ........... 2.2

2.2 Matrix HydraulicConductivityCurve and RelativeHydraulicConductivityand SaturationCurves forUnit PTn ........................ ..... 2.6

'4

2.3 Matrix HydraulicConductivityCurveand RelativeHydraulicConductivityand SaturationCurvesforUnit Chnv ............................ 2.7

2,4 CompositeHydraulicConductivityCurve and RelativeHydraulicConductivityand SaturationCurvesforUnit TCw ........................ , . , . 2.8

2.5 CompositeHydraulicConductivityCurve and RelativeHydraulicConductivityand SaturationCurvesforUnit TSwl ............................ 2.9

2.6 CompositeHydraulicConductivityCurve and RelativeHydraulicConductivityand SaturationCurvesforUnit TSw2-3 ........................... 2.10

2.7 HydraulicConductivityCurve RepresentingShaftsand Driftsand RelativeHydraulicConductivityand SaturationCurvesRepresentingShafts and Drifts in Unit TSw2-3 .......... 2.11

4.1 SaturationProfileand PressureDistributionfrom theCenter of the Model Domainfor the Case ISimulation ...... 4 3

4.2 SaturationProfileand PressureDistributionfrom theCenter of the Model Domainfor the Case 2 Simulation ...... 4 4

4.3 SaturationProfileand PressureDistributionThroughtheLine ElementRepresentingES-I for the Case 2 Simulation .... 4 5

4.4 VerticalCross Sectionof Darcy VelocityVectorsThroughPlane IntersectingES-I for the Case 2 Simulation........ 4 6

4..5 HorizontalCross Sectionof Darcy VelocityVectorsThroughtheLower Part of Unit TCw for the Case I Simulation ........ 4 8

4.6 HorizontalCross Sectionof Darcy VelocityVectorsThroughtheLower Part of Unit TCw for the Case 2 Simulation ........ 4 8

lilt_B '_....... _' _Ip_ '" ,........ ,,,,_,,.... _,_llr,_,'rrlqrli11_Ir''H .... q'' " 'i""' llr"......

4,7 PressureDistributionin HorizontalCross SectionThroughLower Part of Unit TCw for the Case 2 Simulation ...... 4 9

4.8 Saturationsin HorizontalCross SectionThroughLowerPart of Unit TCw for the Case 2 Simulation ...... 4 10

4.9 HorizontalCross Sectionof DarcyVelocityVectorsThroughthe Upper Part of Unit TSw2-3 for the Case I Simulation.. . 4 11

. 4.10 HorizontalCross Sectionof DarcyVelocityVectorsThroughthe Upper Part of Unit TSw2-3 for the Case 2 Simulation. . , 4 11

• 4.11 PressureDistributionin HorizontalCross SectionThroughUpper Part of Unit TSw2-3 for the Case 2 Simulation..... 4 12

4.12 Saturationin HorizontalCross SectionThroughUpper Part of Unit TSw2-3 for the Case 2 Simulation.... 4 13_

4.13 HorizontalCross Sectionof Darcy VelocityVectorsThroughthe PotentialRepositoryHorizonin Unit TSw2-3 for the Case ISimulation ........................... 4,14

4.14 HorizontalCross Sectionof Darcy VelocityVectorsThroughthe PotentialRepositoryHorizonin Unit TSw2-3 for the Case 2Simulation ............ . . .... 4.14

4.15 PressureDistributionin HorizontalCross SectionThrou,gh the PotentialRepositoryHorizon in Unit TSw2-3for the Case 2 Simulation. . . ................. 4.15

4.16 Saturationin HorizonalCross SectionThroughthe PotentialRepositoryHorizon in Unit TSw2-3for the Case 2 Simulation, , , 4.16

4.17 SaturationProfilesfor Unit TCw from the Center ofthe Model Domain for the TransientSimulations ......... 4.16

4.18 SaturationProfilesfor Unit TCw Along the PlaneRepresentingthe Ghost Dance Fault for the TransientSimulations.......... ................. 4.17

4.19 SaturationProfilesThroughthe Line ElementRepresentingES-I in Unit TCw for the TransientSimulations ......... 4.17

xi

, ,i, ....... ,Irl ' rr '111 ,rl I'r qllF' _' ,IIII,_,,j,,,_H,,,,,,,rlll,,,,i,p i, , rl, , '_r'lrrf' ,,,i, llll ' "ll IIp,l_if,lrlP""TI,r' illi , , ' ii,tr 'rl_I .... Irlllr ..... ,.... IIIp,H'

TABLES.

2.1 Fracture Characteristics of Hydrologic Units .......... 2,3

2.2 Parameters for Matrix Hydrologic Properties ........... 2,4i

2.3 Saturated Hydraulic Conductivity Multiplying Factorsfor the Modified Permeability Zone ............... 2.11

2.4 Initial Conditions for the Case i Steady-State Simulation .... 2.13

xii

rip ....... q' iir.... H17 , ,m 'lq ' _P q_' ,' Pl " "q I_ HIll II '.'",', , I1 II qiP' q_il'_' ai' 'q_lN '_ ' rq I,Illaa I1 H,r_HIpr]qr, i I'l'H '1M " lP ' lr' HI'IIH,' _QJq' "'

1.0 INTRODUCTION

Thegeologic formations in the unsaturated zone underlying Yucca

Mountain, Nevada, are being evaluated by the U.S. Department of Energy as the

potential location for a high-level nuclear waste repository. Investigations

are currently under way to evaluate the hydrologic conditions, processes, and

properties of the unsaturated zone at this site.

, Yucca Mountain consists of a series of north-trending fault-block ridges

composed of volcanic ash-flow and ash-fall tuffs that generally have a

regional dip of 5 to 7 degrees to the east (Scott and Bonk 1984). Someof

the welded tuff units are highly fractured, which requires that the hydrol-

ogic properties of these units be evaluated before estimating the rate at

which radionuclides could migrate to the accessible environment. Determining

the significance of fracture flow at Yucca Mountain is important because the

estimated continuum-saturated hydraulic conductivlties of the welded tuff

units are up to three orders of magnitude greater than the saturated hydrau-

lic conductivities of the matrix in the welded tuffs (Peters and Klavetter

!98_). Therefore, if fracture flow is significant, it could greatly reduce

the travel time of water percolating downward through the potential reposi-

tory horizon. The equivalent continuum approximation that was used to

represent the hydraulic properties of the fractured, welded tuffs is

described 'in Section 2.2. The potential repository area, depicted in

Figure i.I, is bounded by steeply dipping faults or by fault zones, and is

transected by a few normal faults. Therefore, the effects of fault zones on

unsaturated water flow and contaminant transport should be investigated.

The exploratory shaft facility (ESF) is an important part of the

subsurface-based site characterization activities that are outlined in the

Yucca Mountain site characterization plan (DOE1988, 1989)o The ESF will

consist of surface facilities and underground excavations, as conceptualized

in Figure 1.2. The underground excavations include two 4.4-m-diameter shafts

(called ES-I and ES-2) connected by drifts at the potential repository hori-

zon. Descriptions of 34 test activities to be performed in the ESF are,,

provided in thesite characterization plan. While the construction and

operation methods will be selected to minimize rock disturbances, some

I.I

,rl_,,i'ill nlFU" , n,,',,ll...... N_l,,rllr,ill_ i 7_,_llU,rl',rllrUl"rlqr' ,'r, ,'lqqr,_r,_r,,_l,,,pm, iiilql,,_rl,,lr,_qnp,r....fql_irlr,,,,_,,,p,_,,rl,l,_r_rq, ,_......m_,l,rr,.,iillUrp,,_i_r, , , ,Irl,,_r,,,...... I' Imll!qIBrlllllllI ,l,l'l...... ,,,,'

CONTOUR INTERVAL 20 FEET

I

I

t "I: _ F ....... "7 "

\ i .... ,.> NAFR L

_, ,,. I

j , L, ....,, .-_ "1 Ii NTS i I I-,--i

J i i i=

. LAS VEGAS '

\, i _.. I <:

_,\ (' -'-,' .....

N765000 - ' 1 _'P4"_

' ,)

'\. !

\ k

•\. '_

(t ,o_o ,o25 0 25 50 Miles

it # Reference Location of

Proposed Repository

CONTOURINTERVAL100FEETN755000- I I

E560_ E565000

YUCCA MOUNTAIN SITE

FIGURE 1.1. Location Map of the Yucca Mountain Site, Nevada(after Lin et al. ]986)

alterations in properties of the rocks close to the excavations are to be

expected. Therefore, it is important to determine the potential effects of

these altered-rock zones [also referred to as modified permeability zones

(MPZs)] on the moisture distribution in the vicinity of the ESF.

1.2

This study is limited to investigation of extremely long-term (steady-

state) and short-term (on the order of 20 years) impacts of the ESF on the

moisture distribution in the unsaturated zone. Long-term effects are impor-

tant for estimating the migration and fate of wastes, while short-term

1.3

Ii, ,q.... rl Ill rf,' ', rl,,,,N .... ,I " llqm'',Ifr "+I' ',',....... _,ir,','r,.... +'_ . , ...... ,rl_," ,_, " 11' ql_, ' ' III..... III ' _ll'll'l_ 11'PTF"_ _,r ',, ,,,, ,',i,i .... ,IPl' '_ ..... Pl ,_,H = r,ll[ ,,ppll_l,,_iil,,ip,m, rl , rf,l,

effects may be important in the planning and interpretation of the tests

performed in the ESF. This study is focused primarily on short-term effectsof the ESF on the unsaturated zone.

,

To study the spatial effects of the ESF, a three-dimensional geometry

was used in the simulations of the unsaturated zone. The exploratory shafts

and drifts, and MPZsassociated with them, are represented as one-dimensional

line elements that are embedded in the general three-dimensional elements.

The Ghost Dance fault, one of the prominent structural and hydrologic

features, is included in the calculation domain and is represented with two-

dimensional planar elements. The ability to incorporate one- and two-

dimensional subdomain features within the general three-dimensional elements

is a unique feature of PORFLO-3(Sagar and Runchal 1990). Results of both

steady-state and transient simulations are presented.

This report is organized as follows. Section 2.0 describes the con-

ceptual model, hydrologic properties, and initial and boundary cond.itions.

Section 3.0 provides a brief description of the PORFLO-3code and options

used for model simulations. Section 4.0 describes the simulation results.

Concl.usions and recommendations for future work are provided in Section 5.0.

Cited references are listed in Section 6.0.

This work was conducted by Pacific Northwest Laboratory( a) for the U.S.

Department of Energy, Office of Civilian Radioactive Waste Management.

(a) Pacific Northwest Laboratory is operated for the U.S. Department ofEnergy by Battelle Memorial Institute.

1.4

"llllIl I ...... _l_l lI II l_ li rl+ 'III1''n+''ll Ip" r' ,ll,lrl IP ..... ,iPl,,riI,t,, _ll+_+ir,T111111rl,Pl', 'II II+ItIPR'Irllll'IIII ...... +I'll1I' _r,r ,' rs111PllP ''"lllm I _l+__ _1111111'+'Hlllr IPIIP

2.0 _.ODELDESCRIPTION

2.1 GEOMETRYOF CALCULATIONDOMAIN

For the purpose of these simulations, the lithologic units at Yucca

Mountain have been grouped into five hydrologic units: I) the Tiva Canyon

. welded unit (TCw); 2) the Paintbrush Tuff nonwelded unit (PTn); 3) the

Topopah Springs welded unit I (TSwl); 4) the Topopah Springs welded units 2

. and 3, which are treated as a single composite unit (TSw2-3); and 5) the

Calico Hills nonwelded vitric unit (CHnv). The thicknesses of these units

are 25, 40, 130, 205, and 130 m, respectively. These thicknesses are based

on the stratigraphy of borehole USW-G4,which is located near the eastern

edge of the potential repository boundary.

The decision to treat TSw2 and TSw3 as a single unit (TSw2-3) is based

on the similarity of the hydrologic properties of the two units and the

relatively small thickness of TSw3 (15 m) relative to that of TSw2 (190 m).

The potential repository is to be located within unit TSw2-3 at a depth of

approximately 310 m below the ground surface (with reference to borehole

USW-G4). i

Computations were performed in a three-dimensional cartesian grid sys-

tem. The axes of this system are aligned with east-west (X-axis), north-

south (Y-axis), and vertical (Z-axis). The dimensions of the calculation

domain are 615, 300, and 530 m in the X-, Y-, and Z-directions, respectively.

A total of 25,056 nodes (29, 16, and 54 in the X-, Y-, and Z-directions,

respectively) were used to represent this domain for numerical calculations

(Figure 2.1). The Ghost Dance fault, which is located near the west bound-

ary, is represented with vertical planar elements (two dimensional), extend-

ing from the ground surface to the water table. The exploratory shafts (ES-I

and ES-2) are represented as vertical line elements (one dimensional),

extending from the ground surface to a depth of approximately 310 m (i.e., inv

unit TSw2-3). The shaft elements are connected by offset horizontal line

elements, representing connecting drifts, at the 310-m depth. The explora-

tory shafts and drifts have a 4.4-m diameter and are assumed to have zones of

modified permeability around them, which extend I m beyond the edges of the

2.1

|

shafts and drifts. Case and Kelsall (1987) report the value of I m as an

upper bound estimate of the expected MPZ dimension for fractured, welded

tuff.

The shafts and fault are positioned within the model domain with refer-

ence to borehole USW-G4 (Scott and Bonk 1984; Fernandez et al. 1988). The

geometry of the computational grid was designed in consideration of the

spatial position of the exploratory shafts relative to the Ghost Dance fault.

The line of symmetry dividing the north and south halves of the model domain,

indicated by the finer spaced grid,, roughly corresponds with the position of

2.2

'I , '.... I ......... ,t '_1' _1 .... _ "' "q_l'l II...... ,I ..... irl_l'_i' '|1' _"_11_'' '" _1_I....... Ilml'l_ ....... I_ " _II '"tip

Coyote Wash, which is an east-west-trending watershed that drains off Yucca

Ridge, located to the west. Tile intersection of Coyote Wash with the Ghost

Dance Fault is a potential location for preferential local recharge, The

finer grid spacing could be used to represent, preferential recharge in futureJ

simulations.

2.2 HYDROLOGICpROPERTIES

The welded tuff units (TCw, TSwl, and TSw2-3) are distinguished hydro-

logically from the nonwelded units (PTn and CHnv) by their matrix hydraulic

conductivities and observed fracture density. Table 2 1 lists the fracture

characteristics of the five tuff units.

Although thera are no direct measurements of fracture properties, the

hydrologic properties of the rock matrix and the fractures are expected to

differ considerably, with respect to both saturated and unsaturated condi-

tions. The saturated hydraulic conductivities for the bulk rock containing

TABLE 2.]. Fracture Characteristics of Hydrologic Units(after Peters et al. 1986)

Fracture Fracture Fracture

Sampl_ Aperture Density{ b) Fractu_ Cor,ductivityUnit Code ) _micronsl (no./m 3) ...... Porosit c) (m/s)

TCw G4-2F 6.74 20 1.4E-4 3.8E-05

PTn G4-3F 27.00 I 2.7E-5 6.1E-04

TSwl G4-2F 5.13 8 4.1E-5 2.2E-05

TSw2-3 G4-2F 4.55 40 1.8E-4 1.7E-05

CHnv G4-4F 15.50 3 4.6E-5 2.0E-04

(a) Fracture properties were measured on a limited number of samples• using confined fracture-testing methods described by Peters et al.

(1984). The differences in the fracture properties of the threewelded units (TCw, TSwl, TSw2-3) reflect different confining

• pressures representing overburden weight evaluated at the averageunit depth in borehole USW-G4.

(b) Based on a report by Scott et al. (1983).(c) Fractures are assumed to be planar; therefore, fracture porosity is

calculated as fracture volume (aperture times I mz) times number offractures per cubic meter.

2.3

L

i

i_ ',l, ........ rl,,,Wlr,, 'r_ I,,_"l 'lrll, ']P "_ql_.... Itri Ill,,Ipl,'lF ilal IIFOi,, ,ra1 .... Nrl,li ,, _,,,llll_,' ,qmll, ,FI_, _ .... ,'ra' ",ali ,_ ' ,lllrln" ,rl rar

fractures are expected to be several orders of magnitude higher than those

of the matrix. Unsaturated hydraulic properties are specified as moisture-

retention characteristics, which describe the relationship of saturation (or

moisture content) with pressure head, and as unsaturated hydraulic conductiv-

ity, which describes'the relationship of hydraulic conductivity with either

pressure head or saturation. For application in numerical models, these

relationships are described by analytical functions (van Genuchten 1978;

Mualem 1976). The parameters of the water-retention and hydraulic-

conductivity functions used to represent the rock matrix of the five

hydrologic units are displayed in Table 2.2 (after Peters et al. 1986). The

corresponding parameters for the fractures are estimated by Wang and

Narasimhan (1986) to be Sr = 0.0395, _ = 1.2851/m, and n = 4.23. These

fracture parameters were used to generate equivalent continuum approximations

of hydrologic properties to represent the fractures in all three welded

units.

The van Genuchten (1978) water-retention curve is defined by

S = (Ss - Sr) [i + (c,_wj)n]-m + Sr

where S = saturation (dimensionless)Sr = residual saturation (dimensionless)Ss = total saturation (=I)

TABLE 2.2. Parameters for Matrix Hydrologic Properties(after Peters et al. 1986)

Grain HydraulicSample Density Total Conductivity

Unit Code (g/cm _ ) Poros i t._ . (m/s,L__ sr _(i/m) n

TCw G4-I 2.49 0.08 9.7E-12 0.002 8.21E-3 1.558 •PTn GU3-7 2.35 0.40 3.9E-07 0.100 1.50E-.2 6.872TSwl G4-6 2.58 0.11 1.9E-11 0.080 5.67E-3 1.798TSw2-3 G4-6 2.58 0.11 1.9E-11 0.080 5.67E-3 1.798 •CHnv GU3-14 2.37 0.46 2.7E-07 0.041 1.60E-0 3.872

2.4

. '" Ii ...... 'vi'l' Pfr...... n itr '_r'lIIl;' "' lr .... Irl,q... Ipi.. rqI1i'' ,'' i, ur"II'lJl

= capillary pressure head (L)

n, m = empirical parameters (dimensionless)

= inverse air-entry pressure head (L'I).

The eff'ective porosity, _e, of each unit is calculated as

_e : ¢_D(I " Sr) (2.2)

where ¢_Dis the total or bulk porosity listed in Table 2,2.

In this analysis, rather than represent the fractures as discrete ele-

ments, an equivalent continuum is stipulated, in which the fractured units

are assigned equivalent or composite properties (Peters and Klavetter 1988).

Explicit treatment of the fractures is not feasible because of the extremely

large number of fractures contained in the site-scale problem and the

difficulties in characterizing and modeling the fracture geometries.

In the equivalent continuum approximation (Peters and Klavetter 1988),

the fracture porosity, _f, can be defined as the ratio of the fracture volume

to the total bulk rock volume. The matrix porosity, ¢_n, can be defined as

the ratio of the volume of the voids in the matrix (excluding fracture void

space) to the volume of the matrix. A bulk, equivalent porosity, ¢_D, can

t llen be defined after Nitao (1988) as

_D : _f + (I - _f)C_n (2.3)

Given the saturation of the fractures, Sf, and matrix, Sm, the equivalent

bulk saturation, Sb, can be defined as

• S_f__Z__+ Sl_._f)¢_ (2.4)Sb : _f + (1 - _f}_n

t

The weighting procedure of Klavetter and Peters (1986) is used to obtain the

equivalent bulk hydraulic conductivity

Kc = Km(1 _f) + Kf_ (2.5)

2.5

Iip' .... '_' I1"' " lPl Ill'rill nif "' ii_, _ li. ,,,i,_ ,,,, ,,irq,i....... I_'lr,, ,,, 11_ II _rr, , qqr" " ,r_' , ' _" '_..... ,11_ I,l':lr. "lt ',,,'_e_ rl' ii 'l,ar,_nl' ,., , ,, Iqpl, , lr ' tri ' "ql"q'"rlrlllrI''llrqq'llll ..... "?lP 'll_P

where Kc, Km, andKf are the composite,matrix,and fractureconductivlties,

_respectively.

Becausethe two nonweldedunits have relativelyfew fractures,their

, composite properties are approximately equivalent to those of the rock

matrix. Thus, for PTn and CHnv, the properties listed in Table 2,2 are used

directly in the van Genuchten (1978) and Mualem (1976) equations. Graphic

representations of the properties of these two units are shown in Figures 2,2

and 2.3. In Figures P..2a and 2.3a, hydraulic conductivity is plotted as a

function of pressure head. In Figures 2.2b and 2.3b, relative hydraulic

conductivity and pressure head are shown as functions of saturation.

Figure 2,4a shows the composite hydraulic conductivity curve for the

TCw unit. A large jump in the value of the composite hydraulic conductivity

1E-5 1ES _. 1

1E-B 1E.5 f K - 1E-2

/rn 1E-7 - 1E-3_E4

E _' 1E.-4

1E-B E 1E-5 ,--._' I 1ooo ,_>'-- _-J 1E-6> 1E-9 U•o I/I .oo 0 100 _ 1E-7 "El

q.) r"lE-lO Z13 1E-8 0

U0 0 10L 1E-B t_

1E-11 _ :>ol 1E-10

,_ ol 0

1E--12 1E-11 oJ

" 1E-12._ 1E-1

_. 1E-13 r 1E-13

tE-14 1E-2 tE-14

1E-15

IE-15 _= 1E-3 - -_--J--, • . I . , I , 1E-18

1_'-3.,_'-_1_'-__ _o _oo1ooo_',, o.o o.,. o.:_ o.B o.8 _.oPressure Head (-m) Saturation •

FIGURE2.2. (a) Matrix Hydraulic Conductivity Curve and (b) Relat, iveHydraulic Conductivity and Saturation Curves for Unit PTn

L

2.6

t_ I ','111 r ".... rll' " II_'lr_l'llH ....... lTI, rltllli,,Fq',, ,,rl_ .... I,,' ,,, m'" ,r' ',_,',,'' ' ,_l' , n,II .... ' ..... I1'1 ''_"" '"'ht"' " "i"'ll_'l= " I'I' IHii,r

1E-5 1Eli . _ 1

_'_ IE-I

ul 1E-7lE+ '

E _ IE-4V E 1E-_ ,_"._ I tooo- _-_ 1E-_B "+'_

+ "I

u o 100 _ . 1E-7 -o-1 q) r-

O"I_ 1E-10 "1- "_ 1E-8C0 _ 10 -_ 1E-9 q)

' 0 1E-11 _ >O ol - . 1E-10 _+_

cn

D 1E-_12 _ I - 1E-11 cD

I - iE-12:>_ t 1E-1

"1" 1E-13 i - 1E-13

' 11E.-14 1E-2 - 1E-14

- 1E-15

1E-15 ..... _ .... _ .... ..i ,,,,_ ..... _ " _ ..... 1E-3 _ I ,,, I , I , I,, - 1E-16

1E-3 1E-2 1E-1 1 10 100 1000 1E4 0,0 0,2 0,4. 0.8 0.8 1,0

Pressure Head (-m) Saturation

FIGURE 2.3. (a) Matrix Hydraulic Conductivity Curve and (b) RelativeHydraulic Conductivity and Saturation Curves for Unit Chnv

at a certain value of pressure head (=-I In in Figure 2.4a) is a distin-

guishing characteristic of the composite curve. In physical terms, the

discontinuity indicates that at a pressure head of -I m (or the corresponding

value of moisture content) flow switches from the matrix-dominated to the

fracture-dominated flow and vice versa. This abrupt change in the hydraulic

conductivity introduces extreme nonlinearity in the problem, which can lead

to numerical instabilities. Figure 2.4b shows the composite relative con-

ductivity and pressure head as functions of saturat, ion for the TCw unit.

, Figures 2.5 and 2.6 show the composite hydrologic properties of the TSwl and

TSw2-3 units. In the numerical model, the discontinuous continuum curves

cannot be easily represented by the analytical expression. Therefore, the

hydrologic properties of the TCw, TSwl, and TSw2-3 units are introduced in

tabular form rather than the analytic van Genuchten relation.

2.7

'['-Pl_l ' ' =rlll ...... II_ .... 1.... lI'11r " 'I" ,'fl ' III ,r, " h' .... I....... IrIHIP..... - nllr' ' "lP' H' r, ' llm, r,lll,h' Irl' 'II, 111..... ,, ,..... "Plllr"%1 I':" +I!

1E-5 1E8 I

IE-I

1E-6 IE5 1F.-2

rf) IE-7 IE-3IE4

E ...--., - 1E-4

--, _E-8 r-.-.,_.-.-.-,,_ E '" _E-_ _-'-P l '°°° r-- "'J -, 1E'-;8 t,j

o I o ,oo ./ -I": 1E--8 0

C rJ

o _ _ lo 1E-gC.) 1E-1 1 :::3 . .

ol 1E-10 '._,L,) ="o. r,.n O

_- 1E-,12"o _ 1E-1::[: 1E-13

1E-14 \ 1E-2 1E-14

1E-1_1E-15 _. ' '"_ ..... -"- 1E-,3 ._.L_._L__..J_...._L_ 1E-16

_E-__-_-i ,i _o 400,iooo1E4 o,o o,_. 0,4 o,B 0,8 _,o

Pressure Head (-m) Saturation

FIGURE 2,4_. (a) Composite Hydraulic Conductivity Curve and (b) RelativeHydraulic Conductivity and Saturation Curves fur Unit TCw

Calculations, based on the continuum properties, are valid assuming that

there are enough fractures of varied orientation to ensure continuum behav-

ior. According to Pruess et al. (1990), the continuum approach will gener-

ally break down for process=.s involving rapid transitions and for conditions

of a very tight Yock matrix or large fracture spacing. These calculations

are also based on the assumption that each hydrogeologic unit is homogeneous,

except where hydrologic properties have been modified for the shafts, drifts,

MPZs, and Ghost Dance fault. Isothermal conditions also are assumed,

although the PORFLO-3 code will solve nonisothermal flow and transport prob-

lems as well.

The air-entry pressure heads (i/_ in the van Genuchten model) for the

MPZs associated with the shafts and drifts in each unit are assumed to be

half as great as the air-entry heads of the rock matrix in each unit. This

2.8

'_" ..... ',, "_p_lr" ""'"'"' IIIIH....... 1 .......... I IIIsl''_....... rl ,_ ..... ,_,,,, H,,,,r.....i,.... ,.,............... 'l_.....Iq'r' '','r_"' IIIll.... FIf_,p,.II,,smrrs....... I,,IIII,..... "111'"' II_FIIII'",'I' ' ''" ..... II,_,,ll,rl'_,

_E 2.5. (a) CompositeHydraulicConductivityCurve and (b) RelativeHydraulicConductivityand SaturationCurvesfor Unit TSwl

modificationis somewhatarbitrarybecausedata ,arenot available,but

reFlectsthe assumptionthat the excavationprocesswill changethe pore-size

distributionin the MPZs such that the MPZs will effectivelybehaveas a

coarserporousmedium than the undisturbedrock matrix. The van Genuchten

model "n" parameter,used to representthe MPZs, was the same as that used to

representthe undisturbedrock matrix, The saturatedhydraulicconductivi-

ties of the MPZs are assumedto be40 and 80 ti:,esgreaterthan those oF the

undisturbedmatrix in units TSwl and TSw2-3,respectively.These multiplying

, factors are based on the upper bound estimates reported by Case and Kelsall

(1987), as shown in Table 2,3. The saturated hydraulic conductivity of the

MPZs in unit TCwwas assumed also to be 80 times greater than t,he matrix

hydraulic conductivity of this unit. This assumption was based on the simi-

larity of fracture densities and porosities between the TCwand TSw2-3

units,as shown in Table 2.1. For the nonweldedunits (PTnand CHnv), the

saturatedhydraulicconductivitiesof the MPZs were assumedto be 20 times

2,9

4

i

iIi 'l_l' ' ' ll""rl I 'PIlp, r, , ', ,ii I ..... +_" "Irl' "'* lll'r ' "qH i11rllll "=_ ,I,P ' Ill Pm'p' ..... II" , i Ill III11 "' IH' '#' ,l lpll'

E.!.GURE2,6, (a) Composite Hydraulic Conductivity Curve and (b) RelativeHydraulic Conductivity and Saturation Curves for Unlt TSw2-3

higher than the hydraulic conductivities of the matrix of each unit. The

nonwelded units have a much higher porosity and lower fracture density than

the welded units. Therefore, it was assumed that permeability modifications,

resulting from shaft construction in these units, would be less extensive

than for the welded units. Ali other properties of the MPZs 'in each unit are

assumed to be equal to the properties of the undisturbed rock matrix in the

units.

The hydraulic properties of the shafts and drifts are represented with

parameters that describe the properties of a coarse material such as gr_vel,

The van Genuchten (1978) model retention curve parameters used to represent

the shafts and drifts are: Sr = 0.0, e = 1.0, and n = 6,2, Figure 2.7 shows

•these curves graphically, These water-retention characteristics were

selected to create a capillary barrier, effectively restricting flow into the

shafts and drifts until the surrounding MPZsare close to saturation. Each

section of the shafts and drifts was assigned a saturated hydraulic

,2,10

II _'IIIl l l'''lll i rll I l lllll l 111111'll l_ lr l -rl ...... -X = _l -l III . II----' _ll_ 'l

TABLE 2.3, Saturated Hydraulic Conduci;iyity Multiplying Factors for theModified Permeability Zone[a) (after Case and Kelsal 1987)

Stress Redistribution ,Without Blast Damaqe Expected( b) Upper Bound(c )

Depth (m) Elastic Elastop]ast:Ic, Case CaseI00 15 2O 2O 4O

310 15 40 20 80

. (a) Equivalent permeability is averaged over an annulus one radiuswide around the 4.4-m-diameter exploratory shaft.

(b) This is based on an elastic analysis with expected strength,in situ stress, sensitivity of permeability Lo stress, and aO.5-m-wide blast-damage zone,

(c) This is based on an elastoplastic analysis with lower boundstrength, upper bound stress, greatest sensitivity ofpermeability to stress, and a l.O-m-wide blast-damage zone.

FIGURE2.7. (a) Hydraulic Conductivity Curve Representing Shafts and Driftsand (b) Relative Flydraulic Conductivity and Saturation CurvesRepresenting Shafts and Drifts in Unit TSw2-3

2.1'I

'1" rlql"l'li .. .. rqpll..,q...lr ._-'-ipH-rR=I -.---.. "'11 ==ll " II q' ' II pl....... _llpq .... I 'r_rll I' ' "l='

conductivity value I000 times greater than the saturated hydraulic conductiv-

ity of the MPZ adjacent to it. Thus, the rate at which water flows through

the shaft and driftelements is governed by the hydraulic conductivity of the

adjacent MPZs.J

No data are available regarding the hydrologic properties of the rub-

bliZed shear zone around the Ghost Dance fault. Therefore, the fault MPZ i

was arbitrarily modeled as 0.5 m wide, with an air-entry pressure head half

as great as the air,entry pressure head of the rock matrix. Saturated

hydraulic conductivities were taken to be 40 times greater than the undis-

turbed matrix saturated hydraulic conductivities of each unit. Tile

van Genuchten (1978) "n" parameters that were used to represent the fault

were the same as those used to represent the matrix in each unit adjacent to

the fault.

2.3 INITIAL CONDITIONS

Two steady-state simulations were performed. In the first, or Case I,

the initial conditions were prescribed as uniform saturation in each unit.

The saturation percentages used were determined from core samples from

various units at Yucca Mountain, reported by Montazer and Wilson (1984).

From these reported average saturation values, pressure heads were calculated

for input as initial conditions to Case I simulations, using the matrix

water-retention curve parameters listed in Table 2.2. These average satura-

tion values and initial pressure heads for the various hydrologic units are

listed in Table 2.4. The shafts, drifts, and MPZ subdomain elements were

excluded from the Case I simulation• The Case I simulation was performed to

establish the initial conditions for Case 2.

The steady-state vertical pressure distribution from the approximate

center of the model domain in the Case I simulation was used as the initial

pressure distribution everywhere in the domain for the Case 2 simulation to

ensure that the specified boundary conditions were consistent with the

internal solution. In the Case 2 simulation, the shafts, drifts, and MPZ

subdomain elements were included in the model domain. The steady-state

TABLE 2+4. Initial Conditions for the Case I Steady-State Simulation(saturation values from Montazer and Wilson 1984)

Unit Averaqe Saturation Initial @(-_ml

TCw 0.67 (0.23) 194s

PTn 0.61 (0.15) 64

TSwl 0.65 (0.19) 232i

TSw2-3 0.65 (0.19) 232

CHnv 0.90 (n/a)(a) 38t

(a) No value given by Montazer and Wilson (1984).

pressure distribution from the center of the model domain in the Case I

simulation was used also as the initial pressure distribution for the

transient simulation.

2.4 BOUNDARYCONDITIONS

The surface boundary condition for the simulations was specified as a

flux boundary, with fluxes representing recharge rates of 0 or 4.0 mm/yr.

The 4.0-mm/yr recharge rate is comparable to the 4.5-mm/yr recharge estimate

cited by Montazer and Wilson (1984). This estimate assumes that after evap-

orative losses, 3% of the average annual precipitation (= 150 mm/yr) at Yucca

Mountain percolates back into the groundwater flow system. A surface bound-

ary flux equivalent to 4.0 mm/yr was applied over a 4-month simulation

period, and alternated with a zero-flux surface boundary condition for an

8-month simulation period for the transient simulation to approximate the

seasonal distribution of precipitation. The lower boundary was specified as

a fixed pressure (atmospheric) boundary, representing the water table.

• The regional dip of the units in the vicinity of the ESF at Yucca

Mountain is approximately 6 degrees to the east (Scott and Bonk 1984).

. Under certain conditions, this may create a driving force in the lateral

(down-dip) direction, which is approximately 10% of gravity. The regional

dip was not explicitly accounted for in the model simulations. However, the

lateral boundaries of the model domain were specified such that a lateral

flow component was created that approximated the effects of the regional dip.

2.13

III_ ..... 11qlql''qlll.... 11 fqp' I .......... ql I? r; II rill ....... :'r;a" ............ :pnl'llllp _Ilqll,r " r:,+,:,,,:rq ..... _q,, IP, 11 ir q,+ " +"r: ,+ I: .... :r.... l;r I:: ,, , IPli'llr..... ,i, ppqll:r1+,iml , ii Irp:" rq':! ' "lP...... rl "qm+i q_.......... "' :llli'' 'I_II rll

The left, or west, boundary of the model domain was specified as a zero-flux

or "no-flow" boundary so that water could not move "up dip" across the Ghost

Dance fault. The north, south, and east boundaries of the model domain were

specified as fixed-pressure boundaries, with pressures corresponding to thei

initial conditions described in the previous section. The east boundary was

specified as a fixed-pressure rather than a zero-flux boundary to create a

lateral flow component, approximating the effects of the regional dip of the

stratigraphy. The combination of the specified boundary conditions creates a

preferential recharge area corresponding to the locations of Coyote Wash and

the Ghost Dance fault, which is consistent with our conceptual model. Future

simulations will utilize a modified version of PORFLO-3, which incorporates a

gravity tensor so that the effects of the lateral driving force created by

tilted beds are directly accounted for.

2.14

i ,, ,,,, ,,,.,,,' _r ' _rlr'rll " ,_ii,,,lll ,,,,, , _, ..... _, ", M, ,.... '"' '' ,"rlilinfl,'"_ml ' Plllr',, i,r ,,iTi i,i '' ',hp .... rl,' qr 'llrll'lql P'I _ Hl'li ' I_,1' _ ' ,,_r 'IIp IIiIPlllll" ,Ip'ltq qpr_'l,"llll lr

3.0 COMPUTERCODEDESCRIPTION

The flow simulations were performed with the PORFLO-3computer code

(Runchal and Sagar 1989; Sagar and Runchal 1990). This code is designed to

simulate fluid flow, heat transfer, and mass transport in variably saturated

porous and fractured media. Up to three-dimensional simulations in eitherI

steady or transient mode can be performed. The steady-state solution can

either be obtained directly by putting the time-dependent term to zero in the

' governing equation or by solving a transient problem until the solution

becomes steady. The first option is very efficient and was used for the

steady-state simulations reported here.

Fractures, faults, shafts, and drifts can be discretized as fully three-

dimensional features or embeddedas one-and two-dimensional features. The

basic assumption for including one- and two-dimensional features in three-

dimensional calculational domains is that the pressure values at grid nodes

represent both the subdomain features within calculational cells (i.e+,

shafts or faults) and the calculational cells (i.e., adjoining rock). Thus

there is no transfer of fluid between the cells and the subdomains within the

cells, as transfer occurs only across cell boundaries. As a result, calcu-

lated pressure values are not as accurate close to the subdomain features as

they would be if the features were discretized as fully three-dimensional

features. Embedding one- and two-dimensional features within a three-

dimensional domain is much more efficient, however, because it requires fewer

nodes and, therefore, less computational time. Because of the large calcu-

lational domain used for the simulations reported here, the shafts, drifts,

and Ghost Dance fault were included as one- and two-dimensional features

rather than being discretized as full three-dimensional features.

• The options available in the code for describing hydraulic properties

of porous media are the Brooks and Corey (1966) or the van Genuchten (1978)

. water-retention models, with hydraulic conductivities determined by either

the Mualem (1976) or the Burdine (1953) formulae. An additional option is to

provide the water-retention and relative hydraulic conductivity data in

tabular form. This last option is useful for specifying the characteristic

curves of a fractured medium that is represented as an equivalent (continuum)

3.1

porousmedium, If tabularvalues are inputfor the hydraulicproperties,the

code 'linearlyinterpolatesbetweenthe values as neededwhile solvingthe

unsaturatedflow equation.

The PORFLO-3code uses the pressure-basedformulationof the flow equa-t

tion that is discretizedusing an integratedfinite-differenceapproach.

Alternatemethodsof solutionare providedto solve the set of algebraic

equations. An iterativeconjugategradientsolutionmethod (Kincaidet al,

1982) is employedfor the resultsreportedhere,

3.2

II i.il_' ,I' ..._,_. r,.,_. .. _i_.rllr. .'_...... Ii' _'_..... rllr', _' ',ft,. ",r, IV pplp.lprq mrl_qr.... ,.... l.i11_llrlql pll_ll ;iHqlrl, , .. ,llr,.ll 11_,_r ' ,q,r li,lql_ .ippr. 1, H11pili, 11'"ri_'l'lllrIr'_ _II,

4,0 RESULTSANDDISCUSSION

The following results represent quasi-steady-state and transient sim-

ulations with upper.-bound MPZproperties (see Table 2,3) at a recharge rate

of 4,0 mm/yr. Other net infiltration estimates fo_ Yucca Mountain range

from 0.1 to 0,5 mtn/yr (Peters et al, 1986), Therefore, the results of thesesimulations should be considered as conservative estimates of the effects of

the exploratory shafts on the hydrology of the Yucca Mountain site.

4.1 STEADY-STATESIMULATIONS

Direct steady-state solutions did not converge at a recharge flux of

4 mm/yr without modifying the hydrologic properties of unit TCw. The primary

cause of the divergence is the abrupt change in composite hydraulic conduc-

tivity of unit TCw, as shown "in Figure 2.4a. With high recharge fluxes,

pressures develop that indicate the beginning of fracture flow (i.e , sat-

uration becomes close to one). However, the recharge is not large enough to

sustain this fracture flow. The result is that the solution alternates from

matrix to fracture flow between iterations and will not converge. Scaling up

the composite saturated hydraulic conductivity of unit TCwby a factor of 25

for tl_e Case I simulation was suf'ficient for successful convergence of the

direct steady-state solution.

By scaling up the saturated hydraulic conductivity of unit TCwfor the

Case I simulation, higher recharge fluxes could pass through unit TCwand

into the underlying PTn unit without jumping from fracture to matrix flow

between iterations, lt is not clear if the solution obtained with this modi-

fication is realistic because fracture flow in unit TCwwas effectively

eliminated. However, the Case I simulation was only used to establish

• initial and boundary conditions consistent with the internal solution at a

recharge rate of 4 mm/yr for the Case 2 simulation, so the absence of pulse

• infiltration resulting from fracture flow was considered to be relatively

insignificant.

The saturated hydraulic conductivities measured on core samples of unit

TCwrange over four orders of magnitude (Peters et al. 1984) In addition to

4.1

l

lt'+I'II+,IHPI''ti_,lqeill,tmllltT' nrl mm,,r,,,,pM ' "'q' ........... " r...... ,,'+_ep,_l'i_'H,'+.... lr, _' hl, p11p,,Ptllil+,_l,,,'leiIpltl+,,,ilr",,,,, ,,_' ,,q'e' "l_lil,,Pm' ' _11' r lr ...... _l__?lll"lllllTllq"Plt ,,qlrp,+," II+'_Tr........ tl1111ll'''"r+,r'

I

this variability,the hydrologicpropertiesof the fracturesreportedby.

Peterset al. (1986)are estimatedrather thanmeasureddirectly (Wang and

Narasimhan1985). Therefore,scalingup tllecompositehydraulicconduc-

tivitiesof unit TCw by a factorof 25 shouldstillyield hydrauliccon-

ductivitieswithin the range of uncertaintyof the equivalentcontinuum

approximations.Withoutthis hydraulicconductivitymodification,steady-

state solutions could not be obtained using the direct solution method, An,

alternative would be to run a transient problem until steady state is

reached. Numerical stability could then be achieved by reducing the size of

the tinl_ steps. However, this alternative is much more expensive computa-

tionally and, therefore, was not used.

A saturation profile through the center of the model domain from the

Case i simulation is shown in Figure 4.1a. The reference elevation for this

and all subsequent figures is the ground surface, "located at a depth of 0 ni.

The shafts, drifts, MPZs, and Ghost Dance fault were excluded from this

simulation. If one could assume that the initial conditions specified for

the Case I simulation represent a quasi-steady state for tile System (in the

absence of tile ESF), and that the hydrologic properties are reasonable, then

the higher saturation obtained in this solution suggest that the long-term

infiltration rate at Yucca Mountain is considerably lower than the 4 mm/yr

assumed here. The pressure distribution corresponding to tile saturation

profile of Figure 4.1a is shown in Figure 4.1b. This pressure distribution

was used for the initial and fixed-pressure boundary conditions specified in

the Case 2 simulation.

As with the Case I simulation, the direct steady-state solution for the

Case 2 simulation did not converge at a recharge flux of 4.0 mm/yr without

modifying the composite saturated hydraulic conductivity of unit TCw. Scal-

ing up the composite saturated hydraulic conductivity of unit TCw by a factor

of 15 was sufficient for successful convergence in the Case 2 simulation. A

smaller hydraulic conductivity scaling factor could be used for the Case 2

simulation because the initial and boundary conditions for this simulation

were more consistent with the internal solution than the uniform average

saturations in each unit that were imposed as boundary conditions in the

Case i simulation.

4.2

II _' " 111'

0

50

lOO ' --'-"-"-t

150

' ' i i

• 200

250

300

350

400

4SOsooI

550 , , -, ..... , ,------,--, : _ , I , , , , , ...._, , , , ._0 0.1 0,2 0.3 0,4 0,5 0.6 0.7 0.8 0.9 1,0 "500 "400 -'300 "200 "100 0

Saturation Total Head(m)

FIGURE 4.1. (a) Saturation Profile and (b) Pressure Distributionfrom the Center of the Model Domain for the Case ISimulation (initial conditions specified for Case 2simulation)

A saturation profile through the center of the model domain for the

Case 2 simulation is shown in Figure 4.2a. The shafts, drifts, MPZs, and

Ghost Dance fault were included in the Case 2 simulation. Comparison of Fig-

ures 4.2a and 4.1a indicates slightly higher saturation in the welded units

for the Case 2 simulation with very little change in the saturations of the

nonwelded units relative to the Case I simulation. The differences in thei

saturation profiles shown in Figures 4.1a and 4.2a show the influence of the

imposed boundary conditions on the internal solution.i

The pressure distribution through the center of the model domain for the

Case 2 simulation is shown in Figure 4.2b. This pressure distribution

corresponds to the saturation profile depicted in Figure 4.2a.

4.3

!Ir ',,r+ r, ,, , " Iri'_,_li r ' 'lli'tll'niu'rql 'irl'l..... lqrlqq ' ,,' " 'iiri" ,,, ,r,i ITill is,rq ii, r, ....... I+I ' ,_'i,_'_r ' ,r;; 'r, ,ql '_ IPll_' '"

FIGURE 4.2. (a) Saturation Profile and (b) Pressure Distributionfrom the Center of the Model Domain for the Case 2Simulation

Also, this figure is provided for comparison with Figure 4.1b, which shows

the pressure distribution from the Case i simulation at the same location.

Figure 4.2 indicates near unit-gradient conditions between the 100- and

350-m depths. If this was a strictly one-dimensional problem, with vertical

downward flow, at steady state one would expect saturations to be nearly

constant between the 100- and 350-m depths. Therefore, Figures 4.1 and 4.2

suggest that a steady-state, one-dimensional flow Field has not been

obtained. However, the solutions are consistent with the specified boundary

conditions.

Saturation and pressure distributions through the line elements, repre-

senting ES-I for the Case 2 simulation, are shown in Figures 4.3a and 4.3b,

respectively, The slight pressure/saturation discontinuity at the interface

lltli " , ,, rl, n ',, pml,',,p,' _, ,, ,, 'P "lql "lllq'l'' ' ' _llt',t * '" rq'qipn',r, ,,ill ,l

o .......... "50

lO0

150'1

. 200]

2s01

aoo]

3501

4oo1 ........ _ ...../

4501

50015501 ! ! ! i i ! i'" ' i ! i '" ! !

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 ,0-500 -400, -300 -200 -100 0

Saturation TotalHead(m)

FIGURE 4.3. (a) Saturation Profile and (b) Pressure DistributionThrough the Line Element Representing ES-1 for theCase 2 Simulation

between units TSwl and TSw2-3 reflects slight differences in the properties

of these units. The maximum saturations obtained from the Case 2 simulation

for units TSwl and TSw2-3 are 0.957 and 0.936, respectively. The hydraulic

conductivity curves shown in Figures 2.7a and 2.7b indicate that water in the

fractures is essentially immobile at saturations of less than about 0.984.

Therefore, the results of this simulation suggest that fracture flow, as

. represented with this equivalent continuum approximation, will not occur in

these welded tuff units.

, In one-dimensional simulations of vertical Flow through the unsaturated

zone at Yucca Mountain using the TOSPACcode, Dudley et al, (1988) used a

2303-node model extending from the water table (0.0 m) to the surface of

Yucca Mourltain (530.4 In). Results of these one-dimensional steady-state

simulations show the TSw unit to be saturated or near saturated, and thus

4,5

I:I' II, ' '11 ip li ' 'ii ', ,iii, rl ..... IH,,IIIFIIr II1"rq....... rtr'_l ' III .... II11 .... illll '1_

amenableto fractureflow at rechargeratesgreaterthan 1,0 mm/yr, This is

contraryto the reSultsof the three-dimensionalPORFLO-3simulations

reportedhere. This differenceis attributedto the natureof tileproblem

and the boundaryconditionsspecified'Forthe three-dlmensionalsimulations,

which allow 'laterallyflowingwater to exit'tilemodel domain, Any lateral

diversionof verticalfluxesabovethe potentialrepositoryhorizondecreases' t

the likelihoodof that horizonreachingsaturatedor near-saturatedcondi-

tions,unless the laterallydivertedwater encountersa preferentialFlow

path such as an MPZ arounda shaft or a fault zone, '

A verticalcross sectionof Darcy velocityvectorsthroughtileplane

that intersectsES-I for the Case 2 simulationis shown in Figure4,4,

Slightperturbationsin the flow field caused by ES-I and the MPZ aroundit

are evidentat the interfacebetweenunits TSwl and TSw2-3 and just abovetile

driftsconnectingthe shafts in i_hepotentialrepositoryhorizonat a depth

of approximately275 m. A potentialfor significantlateralflow above the

interfacebetweenunits PTn and TSwl is indicatedalso in Figure4.4, This

oi! !!ilii........i1111!i!IIII50 ........ _ ..

,,, !iii.iiiiiiii,i. iV V V V V V V V VV V V VVV V V _

250- VVV V V V V V V V V VV V V VVV V Y Y H

VVV V V V V V V V V V V _'VYV V VYV V V Y I

300_V VV V V V V V V V V V V V VV Y V V Y VV V V _( HVVV V V V V V V V V V V V VVVV V V VVY Y Y V 'IVVV V V V V V V V V V V V VVVV V V VVV V V V

0 50 100 150 200 250 300 350 400 450 500 550 600X-Dlstanoe (m)

F1U__. Vertical Cross Section of' Darcy Velocity Vectors Through PlaneIntersectingES-I for the Case 2 Simulation

4.6

lateral f'low 'Is consistent with the results of Rulon et al, (1986) and Wang

and Narasimhan (1988), who also show a potential for significant lateral flow

above tlle potential repository horizon using two-dimensional cross-sectional

flow models,I

A horizontal cross section of Darcy velocity vectors from the Case I

simulation, through the lower part of unit TCw, just abcJvethe interface with

unit PTn, is shown in Figure 4,5, A horizontal cross section of the same

plane depicted in F'!gure 4,5, is shown in Figure 4.6 for the Case 2 simula-

tion, The effects of the exploratory shafts are clearly evident from the

perturbations in the flow field shown in Figure 4,6 relative to Figure 4,5,

The pressure distribution, corresponding to the horizontal flow Field!

j depict(_d In Figure 4,6 for the Case 2 simulation, is shown 'In Figure 4,7.I These pressure contours also "Indicate slight perturbations in the flow Field,

result.lng from the presence of the explore,tory shafts. The saturations,

corresponding to the pressure distribution shown 'In Figure 4.7, are shown in

Figure 4,8,

A horizontal cross section of Darcy velocity vectors f'rom the Case i

simulation, through the upper part of unit Tsw2-3, just below the interface.

with unit TSwl, is shown in Figure 4,9, A horizontal cross section of the

same plane depicted in Figure 4,9 is shown in Figure 4,10 f;or the Case 2

simulation, Again, the effects of the exploratory shafts are clearly evident

from the perturbations 'In the flow "field shown In F_gure 4.10 relative to

Figure 4.9.

The pressure distribution corresponding to the horizontal flow field

depicted in Figure 4.10 For the Case 2 simulation is shown in Figure _.11.

Again, as shown In Figure 4.7, these pressure contours also indicate slight

, perturbations in the flow field, resulting from the presence of the explore.-

tory shafts. The saturations, corresponding to the pressure distribution

shown In Figure 4.1], are shown in Figure 4,12, The boundary conditions,i!

fault placement, and symmetry of this model create a preferential recharge

area on the left (or west) side of the model domain, which corresponds to the

area where Coyote Washcrosses the Ghost Dance fault.

LI,,7

t

11I, II'111",,' 'ml' ' qrrll ..... til '_, _rlll,mlll .... '_'_:lq '" IIIm

_I, _ ..... q,,, '..... n'iilllqr" ,r?itll qp, qirlr ,ii,, ,litr ,Fq, ill ,qtt,l ,lp rf' ' rH.... _I n .1 IIlliilr niiqr"qP ft, ill, _il ',i,r_il ,q_"l,,, ..... Iii' ',,'i *,*, 'iii '_ rllt'"

00 ...... _1 jl i i i J ......... ...... i ..,.. iii J ii i.i i , ,, _ • ..,

.84,.78,- ............. • -_-78,

25O +'

225200

,175 0

,_ 15O E8-2

>L 125 ES-1

lO0

75 _.

&j/j...._

0 _ :t '"'_ " ', ' , : _ r-- ' _, ' i _ m , "'"'i /i,i : T

0 50 100 150 200 250 300 350 400 450 500 550 600

X-Dlstanoe(m)

_FIGURE4.7. Pressure Distribution in Horizontal Cross Section_hrough Lower Part of Unit TCw for the Case 2Simulation

A horizontal cross section of Darcy velocity vectors through the poten-

tial repository horizon, at the 310_m depth in unit TSw2-3 for the Case i

simulation, is shown in Figure 4.13. A horizontal cross section of the same

plane depicted in Figure 4.13 is shown in Figure 4.14 for the Case 2 simu-

lation. The effects of the exploratory shafts are not as evident at this

depth as they are near the interface between units TSwl and TSw2-3 and in

uni t TCw.

, The pressure distribuLion, corresponding to the horizontal -Flow fie'ld

depicted in Figure 4.14, is shown in Figure 4.15, The saturations, corre-

sponding to the pressure distribution shown in Figure 4.15, are shown ini

Figure 4.16, As shown by the flow fields in the previous figures, this

pressure distribution also indicates that the effects of the exploratory

shafts are less significant at the depth of the potential repository than

at shallower depths.

4.9

,,, I'1 _1 m lili'_..... ,,,ll_ ,il,,rilll,II irll....... ,ni .... ,F....... ii_ ', ,_' ,11, ' I'I_" Iii..... _r'"ll lr ......... I I ......... i ......... ii1,11 ii1,......... fill, , _11,,1,_I ................... _1 I, " ....

300 \0.890 0.890 \ ,_,

275 0.895 0,895 .,-0,900- 0,900 ""o

250 '_000

225 c03LC

200 ci Q

'

E 175 i oI15o

o_

a 125>L

lOO0cD ,

75 @ i

0,900 0,900

25 0.890 0,890....

ii_ i ! iI i i , v i i

0 so I00 150 _60 25o _00 3s0 400 4so 5oo s_o 6ooX-Distance(rh)

FIGURE 4.8. Saturations in Horizontal Cross Section ThroughLower Part of Unit TOw for the Case 2 Simulation

Simulations by Wang and Narasimhan (1986), using different constant and

pulse infiltration rates applied to fractures in unit TCw, indicate that most

pulse effects are effectively damped out by units TCw and PTn before water

infiltrates down into unit TSwI. Calculations by Travis et al. (1984) indi-

cate that the penetration distance of pulses of water into fractures con-

tained in densely welded tuffs similar to unit TCw is on the order of i0 m

or less if the fracture apertures are I00 pm or less. Therefore, although

water flow in Fractures may be significant in unit TCw, these simulations

suggest that episodic pulses of water from natural precipitation should not

propagate via fractures into the underlying units due to capillary imbibition

of water into the rock matrix.

4.10

II

3O0tttt t t t t t t t t t t tfttff_ _ j _ ,,,

250 _I_ _ _ I _ _ _ _ _ _ _ _td_l_ _ i .- ...J Al t l i 1 I i I I i I i | Al J d i iii# I i ,_ n,

I::: 200 A Al A A A t A A A A A A A A lA _ _ 4 4 44 4 4 _ _ _."-" AAAA A A A A A A A A l A AAAAdd4444q _ _ ._

" 150 ..........V-_V V 'k_ _ _ ,,,,.,,.,,,.,,,,,,..,,,.,,._ ,,.. ,..,.

,_ V VV Y V V V V V V V V _1 "1 'q'_'_"k_.,,,.,,,.,,.,, ,.,.. ..., _ V VVY V V V V V V V l V _ _ _ __ _ _. _.

_.. 100 vvvv v v v v v v v v v v 't'_,l,_,_,q,_,_,_,_ _ _, ,..I If t t t ! t t t" ! t I I I1111 I _1_,_ _ _ '_ "_

• 50 llll I I I i I I i I i I lllil I I_ _ _ _ "

0 .LE,LI I I I 1 i 1 I _ i I 11111 !ii1!I I v v' ! I' " ! ' li I ! " I

0 50 l C)0 ,,1,50 200 250 300 350 400 450 500 550 600

X-Distance (m)

, FIGURE 4.9. Horizontal Cross Section of Darcy VelocityVectors Through the Upper Part of Unit TSw2-3for tile Case I Simulation

3OO

250 I Jl l l l l l l l A l A 4 A Al A I 4 4 44 A 4 4 4 ..,.^Ak A . A l A A A l A A A A _A A _ _ 144& 4 4 _ ,-

_._ 200 A4_ A l _ l A A l a A _ A _ _ _ _ 4 _44,_ 4 _" ,- --^o A A A A A A 4 _ A A AA A 4 4_,'_¥_.

150 v_ "".-- _,.,_'_ ,_ ._,._.._,. _ _ ,.- ,.

>:.. 100 v,,,_,v v v v v v v v v v I vv,_ ,_,q,_,_,_._.,,. _ ,., ,..V'k_ Y V Y V Y V V V V l I V VVV I I l_q_ _ _ ,_ ,,_

50 Y 1_ Y Y l V V V Y V ¥ Y ¥ V VV Y V I I V_ Y 1 _ _, '_

0 ---ttlt t l t t t V t l t ! lit! I Y _VVY V Y V V' l l II ' '

X-Distance (m)

FIGURE 4.10. Horizontal Cross Section of Darcy VelocityVectors Through the Upper Part of Unit. TSw2-3for the Case 2 Simulation

4.11

FIGURE 4.11. Pressure Distribution in Horizontal Cross SectionThrough Upper Part of Unit TSw2-3 for the Case 2Simulation

4.2 TRANSIENT SIMULATIONS

An objective of this study was to determine the potential effects of the

ESF on the moisture distribution at Yucca Mountain, with emphasis on short-

term characterization work. Therefore, a transient simulation also was per-

formed, using a time-varying recharge rate. The surface boundary flux for

this transient simulation was represented with a step function, so that a

recharge rate equivalent to 4 mm/yr in 4 months (the rainy season) alternated

with no recharge for 8 months. This seasonal recharge was maintained for a

20-year simulation period. This was considered to be an adequate time frame

in which to conduct subsurface characterization work at Yucca Mountain. The

pressure solution from the center of the model domain in the Case ] simula-

tion was used as the initial condition for the transient simulation.

4.12

_i,1!i ........ pr_, ,,, H "' _lllr'q' _"'_"'"..... II1'......... ,,,,' iir,ll ,'pl ......... ,1_ III....'.......'1' "'......"If" ,Irl....."_'_'

300 ............. r , I - _ "

--0,91 0,91- ,

............ _'"_"_"'_275- 0,92- 0,92

'_0250- --_-_--------- 0,93 '0,93---- ,,97,.

225-

200- _'-" O,94

_" 175"

lso-

0,94

,m, a 125-

#.100- _._.-. 0,94-

75- _ + I

50- _----_------ 0,93 2'2225- ,0,92 S

....... 0,91 0,91

O i....... u i u ' I u u' u _'" d.... i i " i-0 50 1O0 150 200 250 300 350 400 450 500 550 600

X-Distance (m)

FIGURE4.12. Saturations in Horizontal Cross SectionThrough Upper Part of Unit TSw2-3 for theCase 2 Simulation '+

No modifications to the saturated hydraulic conductivity of unit TCwwere

necessary for the transient simulation because the size of the time steps was

reduced to achieve numerical stability.

During this simulation period there were no appreciable changes in the

saturations of any of the units other than unit TCw. The saturation profile

for unit TCwfrom the center of the model domain for the transient simulation

is shown in Figure 4.17. The saturation profile for unit TCw, along the

plane representing the Ghost Dance fault for the transient simulation, ism

shown in Figure 4.18. The saturation profile for unit TCw, at the location

of ES-I for the transient simulation, is shown in Figure 4.19. The plot of

time zero in Figures 4.17, 4.18, and 4.19 represents the initial steady-state

solution condition from the center of the model domain in the Case I

simulation.

4.13

II +'+n..... pn,_,....... _,r,'punq.... _rll" pl,i .... qr,lr_ .......n_,rp .....Ip_",'irl,q_...... p, Ii",+,_,!lIlll,,,m_*,"llrnI_r,,'1_'_" ll,,Piiih rlli,,rllf ,_,',,H I_l_lI',_..... I111"_r'" 'llrll....._r i11_.,,i,,'_p+ll_.... , ..... .fr ,,

:300 "i'ttt t t t t t T ' Y f "f t"tff"f ft ffdd 4 _ _ 4

25o tttl t l t l t t 1 1 I I III1 d d _dll i _ .,, ...

i Ai i I l i l i I I i 4 J j lj .1A _ AA44 4 4 .. .

E 200 _ AA A _ _ _ _ _ _ _ A A A _ A_ _ _ A J444 4 4 p.

c_ A AA A A A A A A A A A 4 '_dA A,i 4 ,t # 4 4 4 4 4 _. ,.. .,-

15o .-', "I 4 i "'" " "......

" 0 1111 I I I 1 1 1 I _ I I lllllllttl t t v v! " T-"---- I ' ! ' i l'

o 5'o 'ocX-Distance (m)

FIGURE 4.13. Horizontal Cross Section of: Darcy VelocityVectors Through the Potential Repository Horizonin Unit TSw2-3 for the Case 1 Simulation

300 - '

250 I A_t A 1 l l A i A l i A I A A,A,A ,4 ,4 A AA A I ,t 4 ..

A AJ A _ A, A 'A A A A A A A _ _ _ i1 _ A A.4A 4 4 .. .-

200 AA_X X _ A A A A _ A A A AX_X _ _ _4_4 _..-.._ lm,.

A EA A A _ _ A A A A A A A A AA _ A A A&44 _" ." _" ,-

_L 100 vv_x _ v v v v v v v v v _t_,_,_,q,_,_,_,_,, .,, ,,. .._Y Yl Y V V V V V V Y V V V Y YY'_ I I _'A'A'A % _, _, _.

50 vv_v v _r v v v Y v v v v vv,rv v _ ,_x_,_'t '_ '_ .,,.

0 Hl t t ,.t t t t t t t t t ttr t t t vvvv v v v v, , , , , _) ' (_ _ , , ,0 50 100 150 200 250 3 0 350 4 0 4 0 500 550 600

X-Distance (m)

FIGURE 4.14. Horizontal Cross Section of Darcy VelocityVectors Through the Potential Repository Horizonin Unit TSw2-3 for the Case 2 Simulation

4.14

_r,l ,I n ,'r ",tilt ,," ,,' , irplil,,_q_......... n,, '"rll" ,e,,, ,,,., ,,,,- _p_lv , n.... ml,., r, ,,,,q,ll,lrlIll,,*, "I#_, ll']r'll' 'IRlIP , 'IPl' rl I II '" ii,11 'llP" " ,e....... rf,tmr' nl ,n_,,, ' nl,,,l,ln, ' ,., ""qll' ..... _q,"' ,II " ' ql' IPl_Irll'Irl r l'mir'ele Inlt' '_IqlITP=_"llr ,x,,

300'i

-410 410 +275- .-. 405 +- 405 -- _"_

' "4os. \, \25o------_ -4oo -- ---4oo--_._ --'_ ,'.o

+.oo + -.-. -ago_._. %. \ /|

' '1+ 11//-; +oo__150'

125- ___..._...,._.._,._._.. 390100-

,o..... ,oo :- -405- 40525- 410. +410-

, ,,, , ,........0"' I I I I 1 I 1 I I I I

0 50 100 150 200 250 300 350 400 450 500 550 600X-Distance (m)

FIGURE4.15. Pressure Distribution in Horizontal Cross SectionThrough the Potential Repository Horizon in UnitTSw2-3 'For the Case 2 Simulation

Comparisons of Figures 4.17, 4.18, and 4.19 indicate that, with the

material properties used for these Simulations, the MPZsaY'oundthe explora-

tory shafts create a preferential flow path for water percolating down from

the surface+ The Ghost Dance fault also acts as a preferential flow path,

but its effects are not as significant with the particular properties chosen

to represent it relative to the MPZs and exploratory shafts. At the satura-

tions shown in these figures, fracture flow, as represented with an equiva-

lent continuum approximation, does occur in unit TCw.J

The transient simulations presented in this report suggest that there

will be no appreciable change in the moisture content distribution below unit

TCwwithin 20 years after the construction of the ESF. These simulations,

however, did not account for possible changes in moisture content that

result from additions of drilling fluid into the system during the excavation

4.15

II,., ,,irl_rlllllllll' .... _' " lllilr m' ")'+'m_p'' 'III PI _..... q'll'",. ,,' r'Irl="'' ' ,r"i_I= ',Ph '+_' " ' rl' , .... l_lgll ' '

3oo- o,86 o,88275-_-- 0,87 0,87----------m \ ,0,88------.---- 0,88

250- . 0,8,9.., _ 9(30

0,89 o m

225-_ ,

200- 0,90-----------

175-

' C}_) I

15o-

125- °

100-

__._._.-------- 0,90"75-

50- I_"--- 0'89 /0,88-_ 0,88

25" 0,87 0,87 _I"0,86 0,86

O- '"1 I "' 1 I I I 1 -"='T I r-- "'T"---I'"

0 50 1O0 150 200 250 300 350 400 450 500 550 600X-Distance (rn)

FIGURE 4.16. Saturation in Horizontal Cross Section Throughthe Potential Repository Horizon in Unit TSw2-3for the Case 2 Simulation

iTime (Days)o 0

120

10 + 365

x 1581o 1826

15 v 3407m 3653. 5234

20 • 5479• 7060I 7305

25 _r--'-- ......... , .......0.82 0,84 0.86 0.88 0.90 0,92 0,94 0.96 0,98 1.00

Saturation

FIGURE 4.17. Saturation Profiles for Unit TCw from the Center ofthe Model Domain for the Transient Simulations

4.16

III' _,,,,,_,,,rllI........... _,,,,',,_ ....n' '_....r,,,",,,_,_',IlI, ,, v,,, " lte',, _'_,,,,, ,,,Irt ," rf,,,m,',H", ,_ ,,',,'rlflqq''"_r'r, ......_"' '_" _rn iii'

5 Time (Days)o 0

• _ 120

10 + 365

x 1581o 1828

15 v 3407u 3653. 5234• 5479

20 • 7060u 7305

as ,0,82 0,84 0,86 0,88 0,90 0,92 0,94 0,96 0,98 1.00

Saturation

FI__GGURE4.,I,8. Saturation Profiles for Unit TCw Along the PlaneRepresenting the Ghost Dance Fault for the TransientSimulations

O5 - Time (Days)o 0

120

10 + 365

x 1581o 1826

15 _ 3407II 3653

..... _' " , , .... r'

. 5234

20 • 5479• 7060• 7305

25 , , , +...... ,0,82 0.84 0.86 0,88 0,90 0,92 0,94 0,96 0,98 1,00

Saturation

FIGURE 4.19. Saturation Profiles Through the Line ElementRepresenting ES-I in Unit TCw for the TransientSimulations

4.17

Ill.................... , ....,..............,,_

process. Work by Buscheck and Nitao (1987), however, suggests that imbibi-

tion of the drilling fluid used during the drill-muck-mining excavation

process will be minimal, Therefore, even though shaft construction will

alter the fracture densities and permeabilities of the tuff units, these

simulation,'s suggest that the moisture contents and distributions in the

vi'cinity of the exploratory shafts below l'Cw will not change significantly

within 20 years, which is a reasonable time frame for subsurface characteri-

zation work.

l'he MPZs around tile exploratory shafts in unit TCwappear to create a

preferential flow path for water when fracture flow occurs irl this unit.

This fracture Flow appears to be dampedwith depth, however, so that no

fracture flow occurs in the underlying units. However, lt must be emphasized

that these conclusions are specific to the Initial and boundary conditions,

discretization, and.material properties used for these simulations, and do

not preclude the possibility of lateral flow intersecting preferential flow

paths that result in locally saturated conditions and fracture flow in the

lower units,

6

4.18

I I, ..... H, , , llr

5.0 C_D_CL.USIONSAN.D..RECOMMENDAT!ONS

Irl this model, the exploratory shafts, .drifts, and Ghost Dance fault

are embedded as one- and two-dimensional features within three-dimensional

calculational elements, These featl.ares are'assumed to have tlleir own dis-

tinct .properties, but the results obtained from this model are not as

' accurate in close proximity to these features as they would be if the fea-

tures.were di screti zed as three-dimensional elements, Considering the lack

' of data regardingpropertiesof the Ghost Dance fault and MPZs and the com-

parativelylarge calculationaldomain of the model, the resultspresented

here are reasonableas best estimatesof worst-caseconditionsthai',might

exist. The size of the caIculationaldomainused in this analysisis con-

siderednecessaryto includethe Ghost Dance fault and the exploratory

shafts,while attemptingto minimizeboundaryeffects. If a more accurate

representationof specificsmall-scalefeaturesis desired,the model domain

should be reducedto a smaller.scaleand computationalcell sizes reduced

accordingly, Largerscale approximatesolutions,such as those presented

here, can readilybe used as input to more finelydiscretizedsmallerscale

models within the domainof the "largermodel. On a sufficientlysmaller

scale, the shaftsmay be Includedas three-dimensionalfeatures.

SeveralImportantconclusionsmay be drawn from these simulation

results. The simulationssuggestthat fractureflow, as approximatedby the

equivalentcontinuummodel (i.e.,when hydrologicpropertiesare represented

as shown in Figures2.5 to 2.7), does not occur'at the level of the potential

repositoryat rechargerates up to 4 mm/yr. Discretefracturesor faults

were not simulated,however. Therefore,these resultsdo not precludethe

possibilityof preferentialflow reachingthe potentialrepositoryhorizon

'throughregionsof major structuralfeatures. The Ghost Dance fault is

'includedin this model, but the hydraulicpropertiesused to describetile

fault zone were estimatedbecauseof a lack of data. These simulationsalso

suggestthat there will be no appreciablechangein the moisture-content

distributionbelowthe TCw unit within 20 years after the constructionof the

ESF. This 20-yeartime frame was consideredto be adequatefor subsurface

characterizationwork.

o .

, ,t

5.1

lII_, _ '11 PI al _ ' ii1 ' rDPlPPq' II ir l,pr%, 'iii' 'li?I ' ',lH' ' ........ , 'rl' "' ""1'II1' 'Irl' "

Previous one-dimensional simulations have indicated that the poi:l_rvt'la'l

reposif, ory horizon will be saturated or nearly saturai',ed and t,hus amenabla to

fracture flow at recharge rates of 1,0 mm/yr (Dudley et al, 1988), This Is

contrary to the results of the three-dimensio'nal simulations reported here,I

This difference can be attributed primarily to the dimensionality of the

problem and the nature of the three-dimensional flow field used irl thisI

study, such that water can exit from the model domain through lateral bound-

aries, The results of these three-dimensional simulations support previousii

two-dimensional modeling studies that show a potential for significant lat.-

eral flow above the potential repository horizon, Lateral diversion of ver-

'1tical fluxes above tile potentia repository horizon decreases the likelihood

of that horizon reaching saturated or near-saturated conditions unless the

laterally diverted water encounters a preferential flow path or structural

feature, such as a fault zone,

Postclosure thermal effects and vapor-phase transport of moisture

caused by heat generated from the potential repository were not considered in

•this study. However, these factors may have a significant effect on reposi-.

tory performance, Therefore, postclosure modeling studies should consider

nonisothermal Flow and vapor-phase transport of' moisture. The MPZs and shaft

backfill material may beconlemore important for future smaller scale mocle'l

simulations when coupled fluid flow, heat transfer, and mass transport are

considered.

lt will be impossible to characterize all of the heterogeneities in the

properties of the tuff units at Yucca Mountain in sufficient detail to pre-,

dict all possible outcomes associated with a particular characterization

activity or repository perfornlance after waste emplacement, However, at a

minimum, future simulations with intermediate- and la_ge-scale multidimen-

sional models of the potential repository at Yucca Mountain should inclucle

the topography of the site and surface expressions of major fractures and

faults to determine areas for preferential local recharge. Then, spatially

nonuniform, time-varying recharge fluxes can be applied as surface boundary

conditions to determine the importance of time and spatial variation of

surface recharge to the Flow fields in the natural system at depth, In

addition, "lt may be possible to use multidimensional stochastic modeling

' 5,2 , '

-iw, , , pl,llll i_ ............. til ...... p_, , qp 11_ .... qf' , , _11.... I,,_p ' PJJl..... lP ........ m_, I,qlql,,qrl _1........ I_e ...... _ _"1111," "H'

approaches to evaluate the effects of the va_'iabi'lity and uncertainty of theI

hydrologic properties of the tuff units on ca'Iculai',ions of repository per-

formanceobjectivecriteria,sLlch as clroundwatertraveltime,

|

I

.... .', 5.3

_'_ 'I'' 1lr" ..... ,,H,,,,, II, ' rl ........ n,,,,l, ' ,, _r .... rl[' ,plP,,rl,q ,, ...... H, ,,w1,,,irm,lq; ' fql_' ,_r_ ,, ........... q," I!(l"lll"r' I1'""

o.o B ,ERENCE,

Brooks, R, H,, and A, r, Corey, 1966, "Properties off Porous Media AffectingF'luid Flow," Proceedings of the Amer, Soc, of Civil Engrs,, J, of theIrrig, and Drainage Div,, IR2, 61-68,

Burdine, N, T. 1953, "Relative Permeability Calculations from Pore-SizeDistribution Data," Trans. AIME 198, 35-42.I,

!Buscheck, T, A,, and J, J, Nitao, 1987, Estimates of theHydrologic Impactof Drilling Water on Core Samples Taken from Partially Saturated Densely

' Welded Tuff, UClD-21294, Lawrence Livermore National Laboratory, Livermore,Cal ifornia.

Case, J, B., and P, C. Kelsall. 1987, Modification of Rock Mass Permea-bility in the Zone Surrounding a Shaft in Fractured, Welded Tuff. SAND86-7001, Sandia National Laboratories, Albuquerque, New Mexico,

Dudley, A, L,, R, R. Peters, J. FI, Gauthier, M. L, Wilson, M. S, Tierney, andE, A, Klavetter, 1988, Total System Performance Assessment Code (TOSPAC),Volume I" Physical and Mathematical Bases, SAND85-O002,Sandia NationalLaboratories, Albuquerque, NewMexico,

Fernandez, J, A., T. E, Hinkebein, atld J. B. Case. 1988. Analyses toEvaluate the Effect of the Exploratory Shaft on Repository Performance ofYucca Mountain, SAND85-0598, Sandia National Laboratories, Albuquerque,New Mexico.

Kincaid, D. R,, J. R. Respess, D. A, Young, and R. G, Grimes. 1982."ALGORITHM586 ITPACK2C: A FORTRANPackage for Solving Large Sparse LinearSystems by Adaptive Accelerated Iterative Methods." ACMTransactions onMathematical Software 8(3), 302-322.

Klavetter, E. A., and R. R. Peters, 1986. Fluid Flow in a Fractured RockMass. SAND85-0855, Sandia National Laboratories, Albuquerque, NewMexico.

Lin, Y. T., M, S. Tierney, M. J. Eatough, J. Foster', M. W. Sharp,D. L. Szklarz, D. L, South, B. C. Whittet, R W. Williams, and D. H. Zeuch.1986, Preliminary Estimates of Groundwater Travel Time and RadionuclideTransport at the Yucca Mountain Repository Site. SAND85-2701, S, Sinnock

, (ed.), Sandia National Laboratories, Albuquerque, NewMexico.

Montazer, P., and W. E. Wilson. 1984. Conceptual Hydrologic Model of Flowin the Unsaturated Zone, Yucca Mountain, Nevada. Water-Resources Investi-I

gations Report, 84-4345, U.S. Geological Survey, Denver, Colorado.

Mualem, Y. 1976. "A NewModel for Predicting the Hydraulic Conductivity ofUnsaturated Porous Materials." Water Resources Research 12(3), 513-522,

6.1

Nitao, J, J, 1988, Numerical Modeling of the Thermal and Hydrological Envi-ronment Around a Nuclear Wast6 Package Using The Equivalent Continuum Approx-imation: Horizontal Emplacement, UCID-21444, Lawrence Livermore NationalLaboratory, l.ivermore, California,

Peters, R. R,, and E, A. Klavetter, 1988, "A Continuum Model For WaterMovement in an Unsaturated Fractured Rock Mass." Water Resources Research24(3), 416-430.

O

Peters, R, R., J, H. Gauthier, arld A, L. Dudley, 1986, The Effect ofPercolation Rate on Water-Travel Time in Deep, Partially Saturated Zones,SAND85-0854, Sandia National Laboratories, Albuquerque, NewMexico,

Peters, R. R., E. A. Klavetter, I. J, Hall, S, C. Blair, P. R, Heller, andG, W. Gee. 1984. Fracture and Matrix CharacteristT'cs of TuffaceousMaterials from Yucca Mountain, Nye County, Nevada. SAND84-1471, SandiaNational Laboratories, Albuquerque, NewMexico,

Pruess, K., Y. W, Tsang, and J. S, Y. Wang. 1990. "On Therrnohydrologic Con-ditions Near High-Level Nuclear Wastes Emplaced in Partially Saturated Frac-tured Tuff, 2, Effective ConLinuum Approximation," Water Resources Research26(6), 1249-1261,

Rulon, J,, G. S, Bodvarsson, and P. Montazer, 1986, Preliminary NumericalSimulations of Groundwater Flow in the Unsaturated Zone, Yucca Mountain,Nevada. LBL-20553, Lawrence Berkeley Laboratory, Berkeley, California.

Runchal, A. K., and B. Sagar. 1989. PORFLO-3: A Mathematical Model forFluid Flow, I/eat and Mass Transport in Variably Saturated Geologic Media,Users Manual, Version 1.0, WHC-EP-O041,Westinghouse HanFord Company,Richland, Washington.

Sagar, B., and A, K. Runchal. 1990. PORFLO-3: A Mathematical Model forFluid Flow, Hea_,, and Mass Transport in Variably Saturated Geologic Media,Theory and Numerical Methods, Version 1.0. WHC-EP-O042,Westinghouse HanfordCompany, Richland, Washington.

Scott, R. B., and J. Bonk. 1984. Preliminary Geologic Map of Yucca Mountainwith Geologic Sections, Nye County, Nevada. U.S. Geological Survey Open-File Report 84-494, scale 1:12,000.

Scott, R. B., R. W. Spengler, S. Drehl, A, R, Lappin, and M. P. Chornack.1983. "Geologic Character of Tuffs in tile Unsaturated Zone at Yucca Moun-tain, Southern Nevada," Role of the Unsaturated Zone in Radioactive andHazardous Waste Disposal. J. Mercer, Po S. C. Rao, and I. W. Marine (eds.)Ann Arbor Science Publishers, Ann Arbor, Michigan, pp. 289.335,

Travis, B. J., S. W. Hodson, H. E. Nuttal, T. L. Cook, and R. S, Rundberg.1984. Preliminary Estimates of Water Flow and Radionuclide Transport inYucca Mountain. LA-UR-84-40 (Rev.), Los Alamos National LaboraLory, LosAlamos, New Mexico.

6.2

U.S. Departmentof Energy (DOE). 1988. ConsultationDraft,Site Charac-terizationPlan, YuccaMountainSite, Nevada Researchand DevelopmentArea,Nevada,VolumeII. DOE/RW-0160,Office of CivilianRadioactiveWasteManagement,Washington,D.C.

U.S. Departmentof Energy (DOE). 1989. Site CharacterizationPlan, YuccaMountainSite, NevadaResearchand DevelopmentArea, Nevada,Volume VIII,Part B. DOE/RW-0199,Office of CivilianRadioactiveWasteManagement,

, Washington,D.C.

van Genuchten,R. 1978. Calculatingthe UnsaturatedHydraulicConductivitywith a New Closed-FormAnalyticalModel Water ResourcesBulletin Princeton

f " ,

UniversityPress, Princeton,New Jersey.

Wang, J. S. Y., and T. N. Narasimh_n. 1985. HydrologicMechanismsGoverningFluid Flow in PartiallySaturated,Fractured,Porous Tuff at YuccaMountain.SAND84-7202,SandiaNationalLaboratories,Albuquerque,New Mexico.

Wang, J. S. Y., and T. N. Narasimhan. 1986. HydrologicMechanismsGoverningPartiallySaturatedFluid Flow in FracturedWeldedUnits and Porous NonweldedUnits at YuccaMountain. SAND86-7114,SandiaNationalLaboratories,Albuquerque,New Mexico.

Wang, J. S. Y., and T. N. Narasimhan. 1988. HydrologicModelingof Verticaland LateralMovementof PartiallySaturatedFluid Flow Near a Fault Zone atYucca Mountain. SAND87-7070,SandiaNationalLaboratories,Albuquerque,New Mexico.

6.3

I "Ilq ,,F +,,, qr_...... r,i'" mp,III,,,IM_, i.... lp' 'IIIIq'_lqlP1rl'_",',,Imlql+,,,'qa_l........... 111" qll pi+, +r,' i," Ippl , ,+,,,+1'lr,+ ',II;..... 1..............

PNL-7474UC-814

DISTRIBUTION

No. of No. of ,,__oi__e__ Copies

OFFSITE I. MillerGolder Associates

2 DOE/Office of Scientific 4104 184th Avenue N.E.and Technical Information Redmond, WA 98052

, J.A. Canepa 2 Lawrence Livermore NationalLos Alamos National Laboratory LaboratoryP.O. Box 1663 University of CaliforniaLos Alamos, NM 87545 P.O. Box 808

Livermore, CA 94550H. A. Dockery Attn: W. J. O'ConnellSandia National Laboratory W. HalseyDivision 6312

Albuquerque, NM 87185 T. Pigford,

Department of NuclearB. Hastings EngineeringTRWEnvironmental Safety University of CaliforniaSystems Berkeley, CA 94720

10306 Eaton PlaceSuite 300 M. ReevesFairfax, VA 22030 INTERA, Inc.

6850 Austin Ctr. Blvd.D. Hoxie Suite 300U. S. Geological Service Austin, TX 78731P.O. Box 25046

Denver Federal Center B. SagarDenver, CO 80225 S.W. Research Institute

622 Culebra Road3 INTERA, Inc. San Antonio, TX 78228

101 Convention Ctr. Dr.

Phase II, Suite 540 2 Sandia National LaboratoryLas Vegas, NV 89193 Division 6313Attn: R. W. Nelson Albuquerque, NM 87185

S. Pawah Attn: R. W. BarnardA. E. Van Luik T.E. Blejwas

Distr.l

'II' _I ,_ lr '," ,, ,r, ',lln ...... nj F.,, fqlTn , _rr _ rl _pr'II nnppll_I ",npwlnlll'In'_nnnl'' lr ' , ,' '_' _' _.l'"l.' Tr ,_pp _,_ ,, ......... iln 'n!l _" lllr ,' '_ "_ llll' iIl='"',,,n_ ,rl

PNL-7474UC-814

No. of ONSITECopies -----

DOERichland Field Office2 Science Applications

International Corporation D.C. Langstaff(SAIC)

101 ConventionCtr. Blvd. 2 WestincIhouseHanfordCompanySuite 450

Las Vegas, NV 89109 D.W. LangfordAttn. A. Brandstetter S.J. Ouderkirk

P. L. Cloke

22 PacificNorthwestLaboratory4 U.S. Departmentof Energy

Yucca MountainSite D.M. ElwoodCharacterizationProject D.W. EngelOffice P.W. Eslinger101 ConventionCenter Drive M.D. FreshleyPhase II, Suite 200 C.T. KincaidP.O. Box 98608 M.R. KreiterLas Vegas, NY 89193-8608 K.M. KrupkaAttn. J. M. Boak W.E. Nichols

J. R. Dyer M.L. Rockhold (4)C. P. Gertz D.L. StrengeJ. J. Lorenz J.E. Tanner

R. E. Williford2 U.S. NuclearRegulatory PublishingCoordination

Commission TechnicalReport Files (5)Divisionof EngineeringSafetyWashington,DC 20555 ROUTINGAttn: D. Codell

S. Coplan NLS-260 R.M. EckerJ. W. FalcoM. J. GrahamR. L. SkaggsP. C. Hays (last)

Distr.2

- _' , .... ,,,l_ll"'lr,'.... rl!I_" 'r...... lpr.....!' ......... ii,l,ll,llIl,lr,lrl,,.r . _,,'r_,,r......ilill,,_qn_,_11'rl,'_llT"l,,rlrr'.',lll'lI

iI , , in,lr I lp ii rlll,,,F1prlFi,r ,, F'_" ,1lr' "_pplr''' _ PI......... ,Im_,,_rll_,r@'_','1_'_,_,_,'Fqq__", ,,,, ' ,p_lq_,lr,,',_,' ',Slql,rl,,rj ,IIIF_III jl,l,i "111" ,J_rl,','P_''1rl