&>.; -A , - f %'

. / _ I

SIMULATIONS OF PARADOX SALT BASIN

Ellen J. Quinn and Peter M. Ornstein

8402060151 620824PDR WASTEWM-16PD

3104.1/EJQ/82/08/09/0- 1 -

Purpose

This report documents the preliminary NRC in-house modeling of a bedded

salt site. The exercise has several purposes:

1) to prepare for receipt of the site characterization report by

analyzing one of the potential salt sites;

2) to gain experience using the salt related options of the SWIFT code;

and

3) to determine the information and level of detail necessary to

realistically model the site.

Background

The Department of Energy is currently investigating several salt deposits

as potential repository horizons. The sites include both salt beds and

salt domes located in Texas, Louisiana, Mississippi and Utah. Site

investigations will be occuring in all locations until receipt of the

Site Characterization Report.

In order to narrow the scope of this preliminary modeling effort, the

staff decided to focus their analysis on the Paradox Basin. The site was

chosen principally because of the level of information available about

the site. At the time this work began, two reports on the Paradox had

just been received by NRC: Permianland: A Field Symposium Guidebook of

the Four Corners Geological Society (D. L. Baars, 1979), and Geology of the

3104.1/EJQ/82/08/09/0- 2 -

Paradox Basin, Rocky Mountain Association of Geologists (DL Wiegand,

1981). This in conjunction with the data information from topographic map

of Paradox area (USGS Topographic Maps) and the Geosciences Data

Base Handbook (Isherwood, 1981) provided the base data necessary for the

modeling exercise. The primary goal of all numerical modeling performed

to date was to simulate the contoured piezometric distribution seen on

maps as compiled by Thackson (1981).

The Paradox Basin, located in southeastern Utah and southwestern

Colorado, is underlain by a sedimentary sequence including thick salt

units (Figure 1). The basin itself is defined by the zero salt line

which is the axis separating areas of salt formation from nondepositional

areas. Within the basin, intrusion of igneous features and faulting

associated with dissolution has modified the distribution of the

sedimentary layers (Figure 2).

Two rivers flow through the basin, the Colorado and the Green. The

Colorado River is one of the principal sources of water for the region.

During preliminary studies sites were being investigated both north and

south of the Colorado River (Figure 3). Recently, however, the preferred

sites have been narrowed to an area near Gibson Dome (ONWI 291). The

region selected for modeling is centered around this smaller area.

Use of this basin also provided a parallel effort to the Sandia site

description of the Permian basin. When the modeling began, data

collection on the Permian basin was in the preliminary stages. Sandia

had not developed the conceptual hydrologic flow field and the boundary

conditions remained uncertain. For this reason Paradox represented a

simpler site for modeling purposes. Most of the values selected for

- 3 -

1140 1120 1100 1080 1060420 ---------------- I

I i WYOMING

-N- L W M I N G B A S I N

it | _COLORADO Z400

NEVADA UTAH O

0

380 0Z

-4-

e .. uRS,''> \ , -'~~~--~ ..

.'9 WA-7 )~&,

_ \ A fautr s | V ~ ~ >-Q----. IMIpOXIAT

\/ r \ VERDURE GRAfEN < \ 9 /

X9LRGABNATINSUC E''--'

FAUL. DSHEDZONE APROPATP125,00SCLIMPSMAAITE193

9~ ~~ 20, 40 60 8MIWLAS, 1964

;/,~~~~~~~~~~~WLIM AN HACMA, 1971;Crez

~~~~~~~(e. Wie,-nd, 1981

ARIZONA | W _~~~~~~~~~~ARDO

4AN ~ ~ ~ ~~~~~~HCortez AN YNT 93

FAULT DASED WHRE APROXIATE,(1:25A 00SA LE MAPES) CASIO,2; 73

0 20 40 60 80 l WILLIAMS, 1964:-- - WILLIAMS AND HACKMAN, 1971;WITKIND AND OTHERS, 1978

-Figure .2..(Ref. Wiegand, 1981)

- 5 -

SALT VALLEYGreen River ,UDY

rescent Junction

B Ci ) t )> ~~~~~Uravan\

STY Y AR ESTUDY AREA

| ~~~~~~~~~Monticello \

N~~

25 S sBI~~~~~~~.3lnding ZI j

ELK RISTUDY AE A

! rav\Durango

A R I Z N A -- o -N EW M EX IC O

/7k,' ~ |9F

;~ ̂ EXPLANATI ON

Approximate location of zerothickness of saline facdes

\(boundary of Paradox Basin Study Region)

a- 1 0 20 30 40 Miles

/7; * I - l g

o 10 20 30 40 km

Figure 3.u(Ref. ONWI 290)

3104.1/EJQ/82/08/09/0- 6 -

material properties will be reviewed following receipt of the Permian

basin site definition. It is expected that the basic set up will be

transferable to the Permian basin.

Hydrostratigraphic Units

These analyses have been done on a three dimensional gridding system.

This allowed both examination of the flow in the plane of each layer and

the potential for interaction between layers.

For the initial exercise, three layers were used. These layers

correspond to the three hydrostratigraphic units outlined by Thackston

(1981). The upper hydrostratigraphic layer corresponds to the Permian

and the upper Pennsylvanian strata (Figure 4, 5). The character of the

beds varies from a fairly productive sandstone aquifer in the Permian to

a tight shale layer, the Honaker Trail, overlying the salt. The

variation in conductivity relates to the facies transitions as well as

the fractures associated with igneous intrusions. For initial modeling,

this layer was assigned a thickness of 1500 feet and a conductivity of .5

ft/day.

The middle hydrostratigraphic layer corresponds to the middle

Pennsylvanian units. These include the lower Hanaker Trail and the

Paradox Formation. The rock types are principally shales, salts and

interbed layers. The layer was assigned a thickness of 1500 feet and a

horizontal conductivity of 10 5 ft/day. The lowest layer corresponds to

the lower Pennsylvanian and Mississippian age formations. The lower

Pennsylvanian units, the Molas and the Pinkerton Trail formations, are

- 7 --7-~~~~~~~~~~~~~~~~~~~~~~~~~~

Erathem System Rock Unit

c Alluvial, Eolian ColluvIaland Glacial Deposits

o z Geyser Creek Fanglomerate

Igneous Rock

_ -

Mesaverde Group

a Mancos ShaleDakota Sandstone

Cedar Mtn. Burro CanyonO Formation Formatlon

C.) Morrison Formation

_ _Bluff Sandstone

o o San Summerville FormationIaa Curtis Formation

Group Entraoa Sandstonen ~~~Carmel Formation

CInVo n No a entao rnston

X ~~Chinle FormationF~~~~~~4 enkopi

LI,0N0wi-J

a.

CC

0

Cc

C:C

.C

~i

0

VVnite Rim iDe Chelly)n Sandstone,

O Organ Rock Shale C uter

s ~~~~~~o m tiona Cedar Mesa SFomndo

U Eleonant Can.T_von iosmation$? 1-algeito Shale'Q

o ' Hona er Trail Formation

'UE 8 Paradox FormationI Pinkerton Trail Formation

Molas Formation

Leadville Limestone(Redwsall oQuivaient)

Ouray Limestone

Upper Elbert MemberElbert Formnation

5-McCrackenSandstone Member

Aneth Formation

Lynch Oolomite

Muav Limestone

Bright Angel Shale

lgnacio Formation (quartzites

Upper Hydrostratigraphic Unit

Middle Hydrostratigraphic Unit

iLower Hydrostratigraphic Unit

C

UiB

_-

U00

cra.'

C

.0

E V0

Basement Complex of Igneous

and Metamodphic Rock.

_ C._

Figure 4.(Ref. ONWI 290)

- 8 -

GeneralizedStratigraphicy Column

3 LayerModel

5 LayerModel

. Cutler.' Group

s .

aHonaker

oTrail FM

,Paradox

,Pinkprtpn

- Molas FM

LeadvilleLimestone

0u

,

Layer 1

Layer 2

Layer 3

Layer 1

Layer 2

Layer 3

Layer 4

Layer 5

Figure 5.

3104.1/EJQ/82/08/09/0 _ 9 _

low permeability units while the Mississippian Leadville Limestone is

moderately transmissive. This layer was given a thickness of 1000 feet

and a conductivity of .5 ft/day.

For brine simulations, the three layer stratigraphy was subdivided into

five layers. The upper layer was divided into two to represent the

Permian and the Pennsylvanian Honaker Trail formation. The thickness and

properties of the middle hydrostratigraphic layer, the Paradox salt,

remained unchanged. The lower layer was also divided into two layers to

represent the Pinkerton Trail and the Mississippian formations. The

units have the following thickness, conductivities and brine

concentrations:

1) Layer 1, 1200 ft, .5 ft/day, 0 brine concentration

2) Layer 2, 300 ft, .5 ft/day, .5 brine concentration

3) Layer 3, 1500 ft. 10-5 ft/day, 1.0 brine concentration

4) Layer 4, 300 ft, .5 ft/day, .8 brine concentration

5) Layer 5, 700 ft, .5 ft/day, .8 brine concentration

Description of Geometry

The gridding system used is shown in Figure 6. The northwest boundary

has been extended past the river to ensure that water is not forced into

that discharge point. The northeast and southwest boundaries correspond

approximately to the zero salt lines. They were designed to incorporate

the potential recharge areas along the Abajo and La Salle Mountains. The

choice of the southeast boundary is more arbitrary and is located inside

the zero salt line.

- 10 -

% Approximate location ot zerox thicKness of saline acies

\b'ounirarv ot Praaox dasin .ituav Reqaorr

3104.1/EJQ/82/08/09/0- 11 -

Wells with constant bottom hole pressure were used to simulate recharge

in the mountains. One well was used in the Lasalles and 4 in the Abayos.

For the preliminary study the upper and lower hydrostratigraphic units

are assumed to have no interconnection. Wells are either completed in

the upper or lower hydrostratigraphic unit.

The river is also simulated using wells with constant bottom hole

pressures. The water level in the river was approximated from the

topographic surface map. Future work will require more accurate measures

of the water elevation along the river. No increase in conductivity was

assumed either along the river or around the intrusives.

The pressures assigned to the boundaries are interpolations of the

contours given by Thackston. The number of control points vary a great

deal from boundary to boundary (Figure 7, 8); aquifer influence functions

were used to simulate these boundaries (Figure 9, 10). The salt layer

because of its low conductivity and lack of continuous flow was assigned

a no flow boundary.

Steady State Analysis

Initial runs were performed as steady state. The contours produced

correspond fairly closely to Thackston's contours (Figure 11 a-f). Flow

in-the upper layers were generally toward the Colorado river from both

the north and the south. One exception is the area, south of the La Salle

Mountains where water is moving radially away from the recharge zone

toward the boundary. This reflects our choice of the boundaries as

interpolations of contours rather than structural breaks to flow.

- 12 -

3a Us. *0*

IjeEXPLANATION

AREA OF PRIMARY FOCUSFOR THIS INVESTIGATION

5200 SPRING ELEVATION

*5579 OST POTENTIOMETRIC LEVEL

-- ELEVATION CONTOUR OFPOTENTIOMETRIC SURFACE,SHORTER CDASHES INDICATELESS CERTAINTY

i I

U T A H

LOCArION MAP

- 10 O 20 am

10 0O 3 km

,0,,.

Figure 7. Potentiometric Surface of Upper Hydrostratiqraphic Unit(Ref. Wiegand, 1981)

- 13 -

W !.c- 4,. ¶.

-N-I1EXPLANATION

AREA OF PRIMARY FOCUSFOR THIS INVESTIGATION

4701 POTENTIOMETRIC SURFACEELEVATION

'4ON5 NO POTENTIOMETRIC SURFACEELEVATION GETERMINABLE

T VERY LOW TO ZERG FORMA-TION FLUID RECOVERED

G GAS RECOVERED

c OIL RECOVERED

s SALT WATER RECOVERED

POTENTIOMETRIC RFACEELEVATION CONTOUR

I !__ U T A H

A ~etOA o n I

LOCATION MA1P

0

0

5 IO I5

O 20

*O km

_ 0 ri

Figure 8. Potentiometric Surface of Lower Hydrostratigraphic Unit

f (Ref. Wiegand, 1981) -

- 14

I-4

I

I

N N\

I~

\

'1

Durangt X |

IDuago

I

I" EXPLANAT,0,. 4

N. A "C cnxj-are icar on Of ero

~- - )tOWflfl P300 asin Sj~vR

i' ' latl0 4

o ? 20 20 : 0 I

9s,

* Pressures Used For Te

agraphic Unit

I

- 15

/IIILI

I

I

oUran 90

-Ic~~~~~nI

* *LA

I =N=

E W M E X 5Z-C--I--.

/

1ra 3x d sin ~ L u

:0 -

=

v I

791 )

,IT)

seef NW 2)

Pressures Used For The

Lowr Hdota.Ui

- 16 -

Green River

2.4 00t75207 no... 1.700 ~ ~ ~ 7.1 9~~.(7U7 0.. bo ...... S" ......*5500055P0C405000*0500005.75 5.os c..........

................ . 5 5 % 5 ss. 545555555*

qs54400040.0000 440004000554550 545550004s045o44

0400000500.-. 04ss 55549%%5555 5550955455555 5o*s040t55450,555555sssS 5 "SS * 040400'"0"007005577050755

"S' 55s '4s5

00590

"'sSS5SS5SSSsS 55 5S55

N^^"^^^^^.e-f.--*^^.o~e.K**..I ...............

I ;11.....................ee

I -be 77 *0* 77 ?777?777'7

77777777 970- 00(0 77777777777

,b* '77 777777 ,777'7'7777777

7777777777^-70^*0440 t7777777''7

_~~~~~~~o? 7? r?''7 77 7777'7?'77' 7 77777?????? 7 77ST~t 7 77777070777

_7* t I*^ r wT--?77tr .... It*s! Itszon .......................

_< / q ~~~~~~~~~~~~~~~~~~~~~7777077- 77777?77??S

77777,77,77.7e7¶7777^ ,7Y

007 7??'? 777777,7''n7?7.-77

77777777 77'?77?'7 777?777777777.77 777'-?71?

7005 -7777 ??070 f 777777.77..7.''7777.7 7 7 o7?7-777?777 ?777777777777777'7-7

Ooooo 7"77 777* 00 77777o7?w 7'7 o7 f -. 777777777777777?70'??7' ,,7777*7 I.5 .777 7.

77. ...............7'(* ........ 77 7?','0'r?

77???77? 777 .7 7l -. ?.

eo ..os. 7777 7^^ ..0 0".0* 777777.7,.,. ...

777f7-7-77777777777777z7 -777?7777

000000 777eeta0*S 7777777' ,??2S77--T--~-t77-71-* 0000 '7777 ,77 .. 77777f7'7777t777-7-

7000007750.0. 777-.77

* 000. .0 077000700000700-00A00070074000,000

Col oradoRiver

wIur.t:4?AL r n P.tuc, _or , , c ' I T'l l9ErttCAL c-rSn zI CK p,*r4F. ro I I 7n OFPE'OrFNr VfPfA4s. 4 4141 -.Q r

0'PAC7E

217, * 33. t43. - ASS.650. . .,* . o4 k, , I. ^F .s 3I .O13C. oS * .200F-3

I1.2(997.01 - .SIAE.47j.516E.isr - I. 7 r., 41.7 3?'')' - ,9~

*2.tsA2roc3ŽIeso .4 -.2|eF;

2.7155.q0 - 2.3A2'-02 .3A ' .e b _ o;.o7r 3 2.5 5F 0'703 o0 I 5n x

3.04

10*vI3 - ' * 'E.,75.II4'.O5 .I 5.77)7.0

Figure Ila.SWIFT Output For Layer 1 Steady State Analysis

---

- 17 -

* PARAOOX ASIN FLON MODEL.DARCY VELOCITY

Green River

�13Z1III

1:

q

9

I

eL-

/* f

A . I I . ../I I I I . I Colorado- _4 _ _ D _ ._

vlfft#11�1 14, k%,

�1 4� t �t I

.II II , I II II

II /II \ \,\I//.1

,4 / , I t- ,--/f

t K vul

A7'N,

- 18 -

Green RiverJ.b OOtS'!A$t '^... X tf 4r "E Vew~ru I.e It O"E nl!*! ^ Ciit Ira A.? = 4%e thnt I

abtVtb'abCsbb sassssssssssb5ss5ssssbsssss~seaVsssbss~l

A sabbbabbb A Abb Ss lbSAiab

- 19 -

PAPADOX ASIN FLCW MODELDARCY VELOCI TY

Green River

I'

1-'i

. I ..

I , / ,J I %

a I I

. = = _

N

I:"0~

h L 4 4

% . I.

%4 a A 4

I I 4 A

. . I.

A . . I

A I P I

* 4 b 4 4 a & 4

4 . . & I a 4 4

* . , , b .

A . . I I . I

4 . . .

. h . I . I .

I I I 7 5 P P A

.

.1

I

I

. I,

Colorado

River

2

1,1.r

1,

rn

Pressure Distribution In Feet

PARADOX eASIN FLOW VCCELPRESSURE -0. DAYS

Green River

_ LQCoIorado

River

_. _ 7

'''

- .

N

. _

Figure 11d.

CRSEC Plots Of Layer 2

- 20 -

N

Green River)n7 0O(5iiiO .f (- l *.. tI zr* ' rer vqftuvxt.vt (,.. O& o t. ~,C~ -tl~w ez > e ...t

. . . . . .* xss sssssss555s 5SSSCS4 SSSSS5c s.('; csSc.9S..I ..

* 5 5 5 5 5 5 % 5 5 5 4 5 5 5 5 S 5 S 5 5 5 5I'I 5 %5 5 5 5 5 5 P 5 5 * 5 ~ 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 s 5 5 ~ 5s I .. .... t 5 5

C C 5 5 5 5 5 5 C t C 5< S C% g sP 5 5 5 5 & 5 5 5 5 5 5 5;5 5 5 c5 5 5 5 5 5 5 5 c t s s ss 5; S 5t55w %s~t555t5cs~5tSss c555555I5ss S 5SSC55555 KS SSS S5S tNEt55515..s.IsSS

. 4 i Ss~~~~ssc~5t5555 -.5 5 s5 ~ 5%s55555c 4 ;55q~55 (5 5

- 21 -

I

i-

iz

!I

1. N

!37

I

,

PARPAOX BASIN FLCW MOELDARCY VELCCITY

Green Ri ver. ~ ~~~~~~ /

Coloradoi River

K \, b1

\ t , * .,, ,,

-,\ / s a 4 4 4 b

,, ^ * , .4. ,

,,--- - -,, . . .-. , 4 .

.'--- _, , - . -, . .

..-

K 3

Pressure Distribution In Feet

PARPR

I

/ N~~~~~~~

C~~~ ,N

I;K -

t

RAOX ASMN FLOW MCCELRESSLA R-- Green RiWfS

Figure llf.

CRSEC Plots of Layer 3

3104.1/EJQ/82/08/09/0- 22 -

Flow in the lower unit does not reflect as strongly the topography and

structures of the area. This is due to the presence of the salt layer

which buffers the effects of the overlying units. The lower unit

reflects a larger flow sytem which extends beyond the limits of the

Paradox Basin (Thackston, 1981). The only exceptions to this pattern are

the areas of interconnection along the structures.

The contours produced did not always correspond to the reported contours.

The principal difference occurred along the intrusives (recharge wells)

where calculated heads were higher than those observed. To correct this,

the number of wells simulating recharge was reduced but the calculated

heads still remained high. This could be further corrected by increasing

the conductivity or changing the well index to reflect higher k values

around the recharge zone. This would be consistent with the tectonic

induced fracturing pressumed in the area. Additional refinement may be

obtained by more precisely correlating the locations of wells in the

grid.

Modeling of Brine

Salt concentrations in the Paradox Basin groundwaters cause density

gradients within the overall system, which in turn impacts the

groundwater flow regime. Inclusion of brine in the SWIFT simulations is

an.1mportant step toward creating a more realistic numerical and

conceptual model. In the five layer system the lower four layers contain

varying concentrations of brine while the top layer is considered to be

fresh.

3104.1/EJQ/82/08/09/0- 23 -

The SWIFT steady state flow model was altered to allow the inclusion of

the brine. The first alteration was done by initializing each of the

five layers to a brine concentration and specifying constant brine

concentrations on all boundaries. The middle layer, representing the

salt (Paradox fm), was assigned initial and boundary brine concentrations

of 1.0. Dissolution of the salt layer was not accounted for since the

dissolution parameter contains time units not compatable to steady state

flow.

Results of this simulation were not satisfactory since the brine

concentrations were not correct. Brine concentrations in all layers were

zero, indicating that salt was purged from the system despite constant

concentration boundary conditions. The mass balance calculated on brine

was approximately 1%. This indicated that considerably more salt was

leaving rather than entering the basin. The problem appears to be the

introduction of fresh water in the recharge wells. The analysis will

need to be rerun with brine introduced into the wells.

The next phase in the study was to modify the previous SWIFT set-up to be

a transient simulation. A transient simulation would allow for inclusion

of a salt dissolution term and would enable the SWIFT treatment of brine

to be monitored. First a non-saline steady-state flow regine was

established to initialize flow, and then a transient solution using

saline boundary conditions and a salt dissolution term for the middle

la-er was implemented.

In the transient runs, brine remained in the system but resulted in a

poor mass balance. When SWIFT was allowed to choose its own time step,

the smallest specified step (0.01 days) was chosen with only slightly

3104.1/EJQ/82/08/09/0

- 24 -

improved results. It should be noted that the mass balances obtained are

consistent with results obtained by Sandia (Finley and Reeves, 1982) and

other SWIFT users performing brine studies (Personal Communications).

Clarification is needed from Sandia on what would constitute an

acceptable value.

Variations on the brine dispersion coefficient were included in an

attempt to correct the poor mass balance. Previously, the dispersivity

of brine was assumed to be zero and thus grossly violated the numerical

criteria for brine transport (Reeves and Cranwell, 1981). In order to

satisfy the criteria, the dispersivity coefficient would need to be at

least one half the length of a grid block, or three miles. It should be

noted, however, that the vertical dimensions of the grid blocks are

approximately an order of magnitude smaller than the horizontal. A

dispersivity coefficient chosen on the basis of the numerical criteria

will result in too much dispersion in the vertical direction.

Simulations were performed where the dispersivity criteria was violated,

and also where it was exceeded. Brine mass balance improved with

increased dispersivity values. Where the dispersivity was equal to one

half the horizontal grid block length (satisfying the criteria), the mass

balance was poor. However, at this value complete mixing occurred in the

top two layers contaminating all of the fresh water, and contaminating

the river wells.

Further evaluation of the above results needs to be performed before the

SWIFT modeling study proceeds. A smaller scale study using smaller grid

blocks may be one way of avoiding the brine transport problems

3104.1/EJQ/82/08/09/0- 25 -

encountered above. Incorporation of different logitudinal and transverse

dispersivity will also be useful.

USGS Code

The USGS Trescott, Pinder, Larson (1979) flow code was employed to check

the SWIFT code. SWIFT requires a very complex data set; the results

obtained from the much simpler USGS code would give confidence in the

analyses using the more complicated methods. Although both are finite

difference codes, the different treatments of boundary conditions and

recharge/discharge locations within the respective codes will probably be

revealed in the results. These differences, should they occur, will be

beneficial in analysis and subsequent data calibration.

A steady-state simulation with the USGS code was performed on the top

layer of the steady-state 3 layer SWIFT simulation. The grid

configuration and hydrologic parameters used were identical to those used

with SWIFT. Boundary conditions and recharge/discharge locations were

set by indicating a constant head over each of the specified elements.

This differs from SWIFT where boundary conditions are set by assigning

constant head values to element edges and recharge/discharge locations

are simulated by injection/production wells. The input used for the

codes is contained in Appendix A.

Reasonably good results were obtained with both codes, however, the USGS

code yielded a closer approximation to the real field data (Figure 12).

Preliminary analysis indicates that the head discrepancies are due

primarily to the SWIFT well index values used at the recharge locations

- 26 -

a0.00

O 2

o 2 7. C OS 7.

O 3 50 0 55 bO 6 bl II 52 56 3 2

7 o 5b 05 o b 58 57 * I

x ; *~~~~~~~~~~~~~~~~~~~0.00

7 2 50 3 S *2 6s 60 5 s. 52 4

o II I

X 2 50 - 53 aS 58 57 Sb 57 5b ' 2

.0 2 52 Ss 5 0 S7 Sb S7 4 2

2 R 50 12 56 i5 57 56 S T 6 3

II R 47 R 53 5S 57 bl b7 b7 00 I S 2

* 2 ~So a 53 5D b6 7 S S S I

2~~~~~~~~~~~~~~~~~~ 22 .2 0.00

* q SO R 53 51 1 53 6 b7 59 I 0

o 7.

S a 50 a 51 b 6 b1 59 S7 57 2 2

2 2

3 R2 7 53 7 7 62 67 67 20 ' 2

I.....................,,,,..,,. ...... ... . . . . . . . .. . . . . . . . 0.00

0.0 1.0 2030 30 00 60.00 50.00 60 0 70.00

OTSTANCE MRIN ORIGIN TN CS CTIC4. 1N " ILE3

T 2

2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7

]~ ~~~~~ R A '4 R R R S

2X 1 '3 50| 57 59 57 bO 2

0 R Ss 5 S7 S1 4

s o a s .\ 5 5 7~~~~~S 5* 5 t R

% 5 O 5\Nb 5 7 S 5b 5b

a s 57 is5 5 5b 5D 50 56 ' I

* 0.00

O '3 °2 0 53 .I Sb 67 53 50 -: 2

,~ ~~~~5 _6 6 S. S. N

o 2~~~~~~~~~~~~~~~- a

0 2

. ,.,,,,,,,,,,,,..............................................................__.-._. _ ........... :.. 0

.0 2..0 '30.0 30.0 00.00 5.0 20.0 70.00

00063324r'11G2 TN I 0700001.. "ILL

FigureOL 12 Helad 1ito n t rom 1 1 0m a o

.iur 20.02 itiuto rm SSsmuajn

3104.1/EJQ/82/08/09/0- 27 -

which is allowing excessive amounts of fluid to be injected for a

specified pressure. Further work needs to be done on isolating the

discrepancies.

Recommendations

1. A decision must be made on the level of effort devoted to modeling

the salt sites. The alternatives are either to model each of the

salt sites on a limited scale or to concentrate on one location.

NRC should attempt to find out how much reliance will be placed on

modeling for selection of a salt site. This will determine how many

sites it is necessary to model.

2. Before conclusions can be reached from this modeling exercise, a

more extensive data base should be used. Once additional data has

been compiled, the model would need to be altered to include: (1)

smaller grid blocks around structures to provide better resolution;

(2) inclusion of more layers to reflect the changes in stratigraphy;

(3) spatially variable conductivity to reflect despositional

sequences and structural deformations; and (4) inclusion of

structures such as Lockhart Basin and the Needles dissolution zone.

3. To improve the model given current data limitations, the Siting

Section would need to review the data and recommend the best

characterization of the material. The Siting Section should also

review the head distribution in the drill holes to aid in making

more reasonable estimates of boundary conditions. The recent

.-

- 28

., z0o ,,.

-YN EXPLANATIONAREA OF PRIMARY FOCUSFOR THIS INVESTIGATIONAPPROXIMATE BOUNDARY OFTHE SALINE FACIES IN THEPARADOX FORMATION; ALSOREFERRED TO AS THE ZEROSALT LINE

FAULT SHOWING RELATIVED DISPLACEMENT AT SELECTED

LOCATIONS, DASHED WHEREINFERRED, SOLID WHERE PRE-SENT IN ONLY THE PRE-SALTSTRATA

FAULT DISPLACING STRATA INTHE SALINE FACiES ANDABOVE

G GIBSON DOME NO I BOREHOLEGD-1

i Ii U T A H !i iI I

i L C I O N M A _

LOCATION MAP

C) 5

;k 10

iS 20 m

20 2 Dm.

., ,~

Fiqure 13.

(Ref. iegand, 1931)

3104.1/EJQ/82/08/09/0- 29 -

receipt of ONWI 290 should provide a better data base for future

model studies.

4. A smaller scale model should be developed to analyze faults and

dissolution areas like Lockhart Basin and Shay Graben (Figure 13).

Various interpretations have been proposed to explain the

interactions in the areas. Modeling may be able to aid in

interpreting these collapse features.

5. The USGS code has been successfully used to simulate flow in the

upper aquifer. Since this code has been shown to be simple to use,

use of this code should be expanded to simulate regional flow in the

lower aquifer as well. Such simulation might identify areas of

hydrologic uncertainties and support assumptions that will be made

on boundary conditions in smaller scale simulations.

6. Current contours of flow in the lower aquifer indicate apparent

movement into the region beneath the Colorado River. If correct,

flow could be moving in a more permeable channel beneath the river,

or migrating upward into the river. Such migration would be

introducing salt into the Colorado. SWIFT should be used to

determine the amount of salt that would be transported into the

river by this mechanism' if upward migration would occur. SWIFT

should also be used to investigate the hydrological effects, if any,

that would be caused by more permeable zone underneath the river. A

comprehensive water balance analysis should be performed to

determine the amount of flow from both units seeping into the river

as well as its salt contribution.

3104.1/EJQ/82/08/09/0 - 30 -

7. The errors associated with the brine mass balance in SWIFT need to

be reviewed to determine the problem. SWIFT must be able to provide

reasonable mass balance to ensure reliable representations of

salinity concentrations.

3104.1/EJQ/82/08/09/0- 31 -

References

1. Baars, D. L., 1979, Permianland, Four Corners Geological Survey, p.186.

2. Finley, N. C. and Reeves, M., 1982, SWIFT Self-Teaching Curriculum,NUREG/CR-1968, U. S. Nuclear Regulatory Commission, Washington, D.C., p. 169.

3. Isherwood, D., 1981, Geoscience Data Base Handbook for Modeling aNuclear Waste Repository, V. 1., NUREG/CR-0912, U. S. NuclearRegulatory Commission, Washington, D. C., p. 315.

4. ONWI-290, 1982, Geologic Characterization Report for the ParadoxBasin Study Region, Utah Study Areas, Battelle Memorial Institute,Office of Nuclear Waste Isolation.

5. ONWI-291, 1982, Paradox Area Characterization Summary and LocationRecommendation Report, Battelle Memorial Institute, Office ofNuclear Waste Isolation.

6. Reeves, M. and Cranwell, R. M., 1981, User's Manual for the SandiaWaste-Isolation Flow and Transport Model (SWIFT), Release 4.81,NUREG/CR-2324, U. S. Nuclear Regulatory Commission, Washington, D.C., p. 145.

7. Thackson, J. W., McCully, B. L., and Preslo, L. M., 1981,Groundwater Circulation in the Western Paradox Basin, Utah, Geologyof the Paradox Basin, Rocky Mountain Association of Geologists,Denver, Colorado, pp. 201-226.

8. Trescott, P. C., Pinder, G. F., and Larson, S. P., 1976, Finite-Difference Model for Aquifer Simulation in Two Dimensions withResults of Numerical Experiments, Techniques of Water-ResourcesInvestigations of the United States Geological Survey, Chapter C1,Book 7, U. S. Geological Survey, Reston, Virginia, p. 116.

9. U. S. G. S. Topographic Maps, NJ 12-2, NJ 12-3, NJ 12-5, NJ 12-6, U.S. Geological Survey, Denver, Colorado.

10. Wiegand, D. L., 1981, Geology of the Paradox Basin, Rocky MountainAssociation of Geologists, Denver, Colorado, p. 285.

- 32 -

Appendix A

Input Deck for GS Code

PARA((lXSAL GSL')PISTET I1,mijtll i

hAlf

13 loLON7 52tR0

I I.

I I

4 uo ,450k i.

, 5 j,

55S5*i

50 U0*

5 0 0.5"555 (I i

50)555.Sol,3,5555.500.5555.

555 5000.5555*

SI00.

5100,s5o .

Lsf)P CHEC

52t'u5

I

oS (5 5

5snn,

5559.

555,

S955,6 0 ,

555

5555.

5559.

555,

555S.5?S 0.bOO)).

2 0I 0I)

0 2

0

HE AD

500.LI 5o .5555.SSS.

45 .) .rr 7 0.C55S 9555,SSS5.ss .5555.q9O(0 .

555 .9555.5955.5555559 5.555q.555.555.,5 0.hI o0.

5sss.

5555.,

555,

5 5 5,

55sS.

5F550.

S5559.

Ms9 00.

!9959 555.

C555S.

r55s5.

4s0

Q000*

r55`

555.

5555*

5,555

9000,

b?00.

LISO0.4 5 1) 0 450n 3.

5555,5555,

470 0.

C5 555.5559.

5555

9555 ,55 55.15555.

9555.5 55 5.S5555.

5555,5555.SSSS,S550,5 55 .

555,(I555 O

0500.4500.

4500.500.,

4R00,S5555555.555S.

500.

5555.6500.5555.b9on,

6p500.5555.hSOO.5hv0,

4 J 00.

,01 1IILES

5 555 .

55 5 9.

S5595.

5555.ww'

5555 .

5555.

57uoi

I

. - I 1 . - I .

-I,

_1, * -1, - -1.I.,

-I .-I. .I.

-1.

-. .-I . -.-I .

.5?50 0,

02

I t)

-I. -1, -1.- ,

- . * , _ I.

.1, .1,I

-. -t.

-I. -1. -.

*I. -I. -1 .

I.

*. -I..I.

-1 ...1,

-1.

- I.

.1. -1.

sr S I I 2 4 .

Input Deck for SIr I

I'ATj - -/ '2

;,,.11), ;

:, I (; . t.~* I

3 , * I * o

I, , 31 k

I t ,J j ,\

I * ) ', 'i C ' ( .

.0 x, ~ ~ .

, I

I . ((I, .j

oI ,n

t) I

I .0*t o

o 1 9 n1. 00

I .

1 2 Ii),.l~i I , (3 31 4-

I -5Rl 2,5;R12.5-

1 2 3

R 17HI -SH-It

RiI -I I

9I -12.9I-lb

(C. .) t * 2 4

I .0(

I '' ' ' 'i

.5

2. (I f b.

3.5

.2

.2

O .n

0 .

,) 0

O 0

0 .1

o' .

(3. (

. O'3 ( .(I

. 2 .0

RI-2i.RI .

I -21RI -21RI -1RI -21RI -21RI -21RI -2,RI -27RI -2A

I

w4:t�-

I

5.

3 3 33

S ,5 .(

4

3 3,I, 9

3.uI I I1

3 .j5..)

I. II 4. ,J

I I

2 I,.0

I /112 i;,',

I I

1 .I 7.

I S ., .

I I

1 I

1 i9,I I

I 0"9*.?

J 752.

I I

I I

I ' 5 21 1 li1IS e,,2

3 .

; 3.5

5 .5

3 3

i 3

3 3

3 3

3 3

3 3

1* 3

t )

i, 'J

I . 1

I' I

l: I

I I

I t

; ,J

I I-

1 1

l II ,U

I ,0

.I

1 Y .4 . U

Ii I 4

I , J

II II

I III *

II III -Ic.

II It17 �

II III n

11 IIjb .'�

I IIlb..?.

II IIIc'i�.

II III 7IG�

I iiI? 1?

IA I?.I �

I Ic-'..4

.'

4)I � I &*

I, I�

I 3'Th*1 1

1 4)4)44

I �'4.,) 4;

I 2�4.I;)

I -) '-' 4*.�? L'

I? �.I I

cl 7.I I II

44 lb.II II

4)e? I*II II

%II II

I IIt-..

II IItVl

II II754,.

II II1*p.4.

II II

II IIt. 1�*

II II'4

5, 5

4. 3

3 i

5 3

i 3

S s

S

*1 3

j 3

3 3

1 *4

3 3

3 3

4 3

3 .3

' 3

3 3

i 4,S 5

S SS S

3 I

I I

I I

1 I

I I

I II I1 1

I I

I I1 I

1' 1

IwAU-,

s il Ii I 1l

I I~~~~~~~~~~~~~~~~~~~~R 2

I I ) ? I I

I f ,r .

L , 2 2 5

. 14 2 ) I

E,

F2 5~~~ 2 2 t 1

177C

19 14 1

I . '1 b 5 I I

7 1 4 i * I 7

R I C l

I 7

-- t. j 5

gu, ~~~~~~~~~~~~~~~~~. .I I ., '- ie-

2 I.E'f5S I, F 25" I .f 2Sb I.)tf 5

*o I, JF?57' I. ; E2 5

II I *. tf 25I; I, 'f 9I I ,UF25

1 4 - I , 'F 2 5I S -I,'

17 i.1, F25I 1 - I vF2S

I c. 2 I 1 -3

2i 2 5 i I -32,s 7 21 7,

S, 5 3 I I -32.1-t7 2I7*

- 3 1 1 .

Z* t Z 217,

7 7 S I I -32nE 7 217,

sk S ~I I -3S

I I t I I '-

1I 11 i i I -%

, 12 . I I $ ,tt 1 43

2. 1S I I

I. 2 n5 ro I 142, t 7 Iui9*1 ,

.. i o ¶ I 3.

&i 7 lf I,

I 4 9 3 I -1

I 1 -I 01 1 o-1l 0 100 1 tI 12.0 s.o P2-I UI I, I I) 1 3 '4130.0 (1:100.0

) J V I

w~-4