Vil/"A'4 A center of excellence in earth sciences and engineering

A Division of Southwest Research Institute"' 6220 Culebra Road • San Antonio, Texas, U.S.A. 78228-5166 (210) 522-5160 • Fax (210) 522-5155

November 27, 2001 Contract No. NRC-02-97-009 Account No. 20.01402.861

U.S. Nuclear Regulatory Commission ATTN: Mrs. Deborah A. DeMarco Two White Flint North 11545 Rockville Pike Mail Stop T8A23 Washington, DC 20555

Subject: Programmatic Review of Poster

Dear Mrs. DeMarco:

The enclosed poster components are being submitted for programmatic review. These poster

components will be assembled as a poster and will be presented at the American Geophysical Union

Fall Meeting to be held December 10-14, 2001, in San Francisco, California. The title of this poster is:

"Monte Carlo Analyses of Unsaturated Flow in Thick Vadose Zones of Layered, Fractured Rocks" by W. Illman and D. Hughson

Please note that this poster is based on an abstract previously submitted to NRC on

September 7, 2001. Please advise me of the results of your programmatic review. Your cooperation

in this matter is appreciated.

Sincerely,

Budhi Sagar Technical Director

/ph Enclosures cc: J. Linehan K. Stablein W. Patrick

B. Meehan D. Brooks CNWRA Directors E. Whitt B. Leslie CNWRA Element Mgrs J. Greeves N. Coleman T. Nagy (SwRI Contracts) J. Piccone P. Maldonado W. Reamer T. Essig

Washington Office • Twinbrook Metro Plaza #210 12300 Twinbrook Parkway - Rockville, Maryland 20852-1606

Monte Carlo Analyses of Unsaturated Flow in Thick Vadose Zones of Layered,

Fractured Rocks Walter A. IIman

University of Iowa (walter-illman@uiowa.edu)

Debra L. Hughson Center for Nuclear Waste Regulatory Analyses (CNWRA)

Now at Mohave National Preserve (DebraHughson @nps.gov)

1

Acknowledgements

* This poster was prepared to document work performed by the Center for Nuclear Waste Regulatory Analyses (CNWRA) for the U.S. Nuclear Regulatory Commission (NRC) under Contract No. NRC-02-97-009.

* The activities reported here were performed on behalf of the NRC Office of Nuclear Material Safety and Safeguards, Division of Waste Management.

• This work is an independent product of the CNWRA and does not necessarily reflect the views or regulatory position of the U.S. Nuclear Regulatory Commission.

2

Abstract We conducted Monte Carlo simulations of flow in unsaturated fractured rocks using a two-phase, nonisothermal, flow simulator. In this simulator the fractured rock is idealized as a dual-continuum porous media, in which the matrix and fracture constitute two distinct continua represented by two overlapping, interacting numerical grids. Darcy's law and the area of the matrix-fracture interface open to flow govern the exchange of fluids between the two continua. To investigate the applicability of the dual-continuum approach for modeling unsaturated flow in a thick vadose zone of fractured rocks, we applied the model to site data collected from Yucca Mountain. We used grid blocks with dimensions of 1 m that is commensurate with the support volume of fracture permeabilities estimated from single-hole pneumatic injection tests. We investigated the consequences of simplifying fracture permeability on unsaturated flow by comparing the model results using uniform formation properties to a stochastic model that represents spatial variability of the fracture permeability within the layers as a multivariate lognormal random field. In both models, the water flux boundary condition was varied to simulate the effects of variable recharge rates.

We found that the variability in fracture permeability causes the development of preferential flow paths in the fracture continuum for the welded tuff units and in the matrix continuum for the nonwelded unit. The magnitude of variance in fracture permeability correlates well with the degree of flow focusing. Water flow rates in preferential flow pathways have been found to be locally very high (more than ten times the input flow rate). Flow focusing due to the development of preferential pathways increases saturation locally. This local increase in saturation causes an increase in relative permeability to water along the pathway and may reduce the wetted surface area for fracture-matrix interaction.

Comparison of results obtained from the homogeneous and heterogeneous models of unsaturated flow through thick vadose zones shows that deep percolation can take place rapidly through persistent, prefe4=nfial flow patThese pAt a s are har o detect and may carr large volumes of water. Simpl ationN sitei geoloqjltto IroneojicluThns e spatial and temporal

3

Study Objectives "* To compare deterministic and stochastic continuum analyses of

water flow in thick vadose zones of fractured rocks.

"* To investigate the effect of flow focusing on deep percolation in

layered, heterogeneous media.

"* To investigate the effects of varying o2 on flow focusing.

"* To investigate the effects of varying water flow rate (q) on flow

focusing.

"* To investigate the consequences of simplifying model structure

and parameters on deep percolation.

4

Current DOE conceptualization of Yucca Mountain

I. ..I ' I'

2000 GFM DEVELOPED BY LBNL

EAST-WEST SECTION B-B' THROUGH SO-7

1 800

S1600 w 0N

zV o 1400 0Z F- o 11

w _j 1200

1 000

800 - r...__ h __p, v

tm 3vp_

1.6IOE5 1.7000E5 1.71 00E5 1 .7200E5 1 .7300E5

EAST NEVADA COORDINATES (i)

5

Complexity of Unsaturated Flow in Thick Vadose Zones of Fractured Rocks

"* Episodic nature of precipitation events

"* Localization of infiltration

"* Poorly understood transport of water, air, and heat in the shallow

subsurface

"* Poorly understood constitutive relationships for unsaturated

fractured rock.

"* Spatial and temporal variability of water and air flow resulting from

heterogeneity in rock properties at multiple scales

"* Difficulty in measuring ambient percolation rates in situ

6

-36 Data in the ESFEvidence for fast flow?

0 20 40 60

ESF station [102 m]

7

Cl

4500

4000

3500

3000

2500

2000

1500

1000

LC)

x

CfJ 03

500

0

80

Simulation of permeability fields0 Permeability fields - direct Fourier

Transform Method developed by A. Gutjahr and his coworkers (Robin et al., 1996).

• 10 unconditional simulations were conducted to generate permeability fields with isotropic correlation scales for the TCw, PTn, and TSw units.

0 Lack of information at Yucca Mountain precludes rigorous geostatistical analysis of data.

Exponential model

y(h)=U'2 i-exp(-hj

h = separation vector magnitude

?= correlation length (2 m)

o2= variance (0.5 - 2.0)

8

Mass and Energy TRAnsport (METRA) Lichtner et al., (2000)

"* Two-phases (fluid and vapor), non-isothermal flow "* Computations in 1-, 2-, or 3-D "* Block-centered structured / unstructured grids: Integrated Finite Volume

(IFV) "• Single, dual, or equivalent continuum "• Flexibility in incorporating various empirical relations to describe two-phase

flow properties (van Genuchten, Brooks-Corey, linear, or user-defined) "* Three primary equations solved:

- Total mass balance - Air mass balance - Energy balance

"* Primary field variables for two-phase problems: - Total gas pressure (Pg) - Partial pressure (Pa) - Gas saturation (sg).

"* Solution based on a fully implicit formulation using a variable substitution approach

o* pat I he,.ceneit rok flow~ad thermal paramLes

9

Model Assumptions 9 2D, isothermal simulation

e Dual continuum model with fractures treated as stochastic continuum; matrix treated as homogeneous continuum.

0 1 m3 grid blocks

* 2 phase flow (air and water)

• Mixed BC at the top; no-flow BC at the sides and constant head BC at bottom.

10

Numerical Simulation Parameters

* DCM parameters from TSPA-VA * Fracture permeability statistical

parameters: Mean values from single-hole pneumatic injection tests at Yucca Mountain.

- Variance and correlation lengths chosen arbitrarily. Variance also derived from geostatistical analysis of air injection tests at the Apache Leap Research Site (ALRS).

11

Case 1 o2= 0.5; q= 42.5 mm/yr

Water Flux in Fracture Fracture continuum, Continuum

10 L103

S•.• -8.8 u~vv ~ ,-9.2 2

S-10.0 30 .. 4 . ..- 10.3

-10.7 40 -11.1 40

-11.4£ 50 -11.8 h1 50~LEv 50

I -12.2 N -12.6

60 -12.9 60 -13.3

70L'_.... -13.7

70 ,.-14.0

-14.4 80 . ~ *-14.8 80

-15.5 .i 90 - • T : -15.9290

4 U nU t x [m]x [m]

12

Case 2

Fracture continuu

41F~

t' 'W

x [m]

1.0; q = 42.5 mm/yr Water Flux in Fracture

Im Continuum

A 10 IIUMW ULOG 10k

-8.8 -9.2 -9.6

-10.0 -10.3 -10.7 -11.1 -11.4 -11.8 -12.2 -12.6 -12.9 -13.3 -13.7 -14.0 -14.4 -14.8 -15.2 -15.5 -15.9

20

30

40

"E 50 N

60

70

80

90

x [m]

10

20

30

40

.50 N

60

70

80

90

13

Case 3 o2= 2.0; q = 42.5 mm/yr

Fracture continuum, Water Flux in Fracture Continuum

10 10

LOG1Ok 20 -8.8 20

-9.2 -9.6

0 -10.0 20 -10.3 -10.7

40 -11.1 40

E -11.4 ,-,50 A--1.850

-12.2

f4-" 4 V " S-15.5

60 -- 12.9 60 ' -13.3

70 """""===="= .... . .. . . ., ,-13.7 S...... :• ,.4~ f M • I' .•L•, dlP• ,/" •% .i -14.0 7

AS• :,• ; ,, -14.8 8

, .. ... •"•-:• -15.5

90 .. ,,i•.••,.: ¢!. Z:,>:%. ;..l. 1•.,i - 15.9 90

50 60 x [m]

10 20 30 40 50 60 70 80 90 100 x [m]

14

Case 1 = 1.0; q = 42.5 mm/yr

a) Saturation in fracture continuum lo

40

60;

70

90

0 1000. . .4 .50 6 70 80 90

b) Saturation in matrix continuum

70

10

c) Water flux in fracture continuum

10

20

30

42O

40

60022907

70

80 1018

10 20 30 40 50 60 70 80 90 100

x [nm]

ch Water flux in matrix continuum

&037230

7I26

x(m]

Observations

"* High variability in saturation of

fracture continuum in welded

units (TCw, TSw).

"* Variability in saturation of matrix

continuum in the nonwelded unit

(PTn).

"• Flow focusing in fracture

continuum of TCw and TSw

"* Flow focusing in matrix

continuum of PTn.

15

•J

o. 92

0 .=

oa7

o71 070 07

°I

Variation in flow rate at top boundaryo2= 0.5; q = 12.5mm/yr 02= 0.5; q = 22.5mm/yr

20

S50

60

184 10 15589 14769 139.46 13128 :2307 11487 1 066

9846 9025 8205 7394

6564 57 43 4923 47102 3282 2441

1641 8020

o2= 0.5; q

10

20

30

40

T.s

= 42.5mm/yr

x([m]

o2= 2.0; q = 12.5mm/yr 02= 2.0; q = 22.5mm/yr c2= 2.0; q = 42.5mm/yr

Uk 40886 38841 367 97 347

0 34753

32649 30664 28620 265 7 245 31

224 87

20443 14399

ý63054

833

2034m

10 20 30 40 50 60 70 80 90 100 x [m)

9c

30

so

218 to 207.19 19629 18338 17443 163 57 15267 141 78 130 8 11995

10905

98 14 8724 7433 8343 5452 4362 3271 21 81 1093

x [m]

90

" 57 N

772 19 7 33

654 34 - 1715

,579 14

540 53

"424 70 930 70

16

N '

Conclusions "• Preferential flow paths develop in the fracture and matrix

despite the uniform application of water at the top boundary and without explicitly building in high permeability pathways or discrete features that represent fractures.

"• The development of preferential pathways in the fracture continuum increases the relative permeability to water along these pathways and reduces the wetted surface area for fracture-matrix interaction.

"• Water flow rates in preferential flow pathways can be locally very high (more than ten times the input flow rate).

17

Conclusions (cont.) The presence of the nonwelded PTn in the UZ led to the past conceptualization that deep percolation at Yucca Mountain becomes attenuated and laterally diverted once it reaches the PTn unit. Our model shows that:

- (a) rapid flow takes place through persistent, preferential flow paths in the TCw and TSw units;

- (b) the focusing of flow at the TCw/PTn contact causes localized increases in matrix saturation that can extend from the TCw/PTn contact to the PTn/TSw contact;

- (c) increase in saturation causes the development of preferential pathways in the PTn matrix continuum; and

- (d) the preferential pathways allow for the rapid, predominantly downward movement of water through the unit.

18

Conclusions (cont.) "* Comparison of results obtained from the homogeneous and

heterogeneous models of unsaturated flow through thick vadose zones shows that deep percolation can take place rapidly through persistent, preferential flow paths.

"* These pathways are hard to detect and may carry large volumes of water. Simplification of site hydrogeology may

lead to erroneous conclusions on the spatial and temporal distribution of unsaturated flow through thick, fractured vadose zones.

"* The effect of episodic infiltration on unsaturated flow through thick, Il d vadose zone is currently under investigation.

19

References "* Civilian Radioactive Waste Management System Management and

Operating Contractor, Chapter 2 Total System Performance AssessmentViability Assessment (TSPA-VA) Analyses Technical Basis Document, Unsaturated Zone Hydrology Model, B00000000-01717-4301-00002 REV 00, Las Vegas, NV, Civilian Radioactive Waste Management System Management and Operating Contractor, 1998.

"• Flint, L. E., 1998, Characterization of hydrogeologic units using matrix properties, Yucca Mountain, Nevada, US Geological Survey Water Resources Investigation Report 98-4243, US Geological Survey, Denver, CO.

"* Lichtner, P. C., M. S. Seth, and S. Painter, 2000, MULTIFLO User's Manual, MULTIFLO Version 1.2, Two-phase Nonisothermal Coupled ThermalHydrologic-Chemical Flow Simulator, Center for Nuclear Waste Regulatory Analyses, San Antonio, Texas, Rev. 2 Change 0, January 2000.

"* Robin, M. J. L., A. L. Gutjahr, E. A. Sudicky, and J. L. Wilson, 1993, Crosscorrelated random field generation with the direct Fourier Transform Method, Water Resources Research, Vol. 29, No. 7, 2385-2397.

20