Hydrogeology of the WIPP SiteRichard L. BeauheimRobert M. Holt

IntroductionThe northern part of the Delaware Basin isfilled with approximately 24,000 ft ofPhanerozoic sedimentary rocks (Hills, 1984).At the WIPP site, the upper 4,000 ft ofsedimentary rocks contain evaporites (Powersand others, 1978) (Figure 1). The nearestmajor aquifer is the Capitan reef, which. islocated about 10 miles from the WIPP sIte(Figure 2). The water-bearing zones withinthe rocks which underlie, host, and overlie theWIPP repository have low permeabilities andstorativities, are generally confined, andcontain waters with high salinities and longresidence times. As part of WIPP sitecharacterization activities, each of the majorwater-bearing zones has been investigated.Several studies continue to supplement dataand refine our understanding of thehydrogeologic system. The major formationsof interest at the WIPP site are discussedbelow, in stratigraphically ascending order.

Bell CanyonFormationThe Bell Canyon Formation is the uppermostformation in the Permian-age DelawareMountain Group (Figure 1). The upper BellCanyon is subdivided into five informalmembers: the Hays sandstone, Oldssandstone, Ford shale, Ramsey sandstone,and Lamar limestone (Figure 3). Individualchannel sandstone units are separated laterallyby stratigraphically equivalent siltstones andshales.

The most significant hydrologic units withinthe Bell Canyon are the channel sandstones.The channel sandstone units trend roughly

northeast/southwest through the basin (Figure4). Intergranular porosity ranges from 20 to28 percent, while hydraulic conductivitiesrange from 3 x 10-2 to 2 x 10-1 ft/day (Berg,1979). The siltstones and shaley siltstoneshave porosities ranging from 10 to 20 percentand hydraulic conductivities less than 2 x 10-4

ft/day (Mercer, 1983). Tests conductedduring WIPP site characterization activities atboreholes AEC-7, AEC-8, and ERDA-I0showed the hydraulic conductivities of BellCanyon sandstones, siltstones, and shales torange from 2 x 10-6 to 5 x 10-2 ft/day (Mercer,1983). Mercer (1983) reports these values tobe consistent with the range of permeabilityvalues reported for the Bell Canyon by variousoil companies.

Within the Bell Canyon, the shallowestsignificant hydrologic unit is the Ramseysandstone. At borehole DOE-2, the mostpermeable portion of the Ramsey appears to beabout 28 ft thick (Beauheim, 1986). TheLamar limestone acts as a confining layer.The Ramsey consists of laterally and verticallyinterlayered sandstones, siltstones, and shalesdeposited as deep-basin channels and relateddeposits (Lappin, 1988).

The Hays sandstone is the most permeable andtransmissive unit within the Bell Canyon nearthe WIPP site based on hydrologic testing intwo other WIPP wells (Cabin Baby-l andDOE-2) (Beauheim and others, 1983;Beauheim, 1986). Bell Canyon hydraulicconductivities were reported by Beauheim andothers (1983) to range from 1 x 10-6 (Lamar)to 4 x 10-3 (Hays) ft/day at Cabin Baby-l andby Beauheim (1986) to be 2 x 10-4 (Ramsey)to 6 x 10- 3 (Hays) ft/day at DOE-2.

-131-

SYSTEM SERIES GROUP FORMATION MEMBER

RECENT RECENT SURFICIAL DEPOSITS

QUATERN- PLEISTO- MESCALERO CALICHEARY CENE GATUNA

TRIASSIC DOCKUM UNDIVIDED

DEWEY LAKERED BEDS

Forty-niner

Magenta Dolomite

RUSTLER Tamarisk

Z Culebra DolomiteC unnamed0~00

SALADO

ZC-~a::

CASTILEw~

ZC(

BELL CANYONZ ...C ziL ::)

0::)~-'

C wQ a:: CHERRYc C::) ~ CANYONCJ C

-'w BRUSHYQ

CANYON

Figure 1. Generalized stratigraphic column of the Delaware Mountain Group and youngersedimentary rocks at and near the WIPP site.

-132-

TEXAS

NEW MEXICO

oU

xl(J)LU <{~ X

~ ~LUZ

Jf

$

N

--"'- / ~'"

~~-..........,A \,Ol-t-\~ \\

\

"\\,,I\,'~JALlr\\\

iIi

B!A SIIIII

o 5 10 mlI ! !

I.........V.JV.J

I

Figure 2. General geological and hydrological setting of the WIPP site relative to the northern Delaware Basin.

SYSTEM SERIES FORMATION MEMBER

Z W< ..J0 -t-::I: enU <0 U

lamar

Z Z

~0 Ramsey>

:E za: < Fordw UQ.

..J

..JZ w Olds~ CDQ.::> Hays..J<C< z::>C} 0

>Z<U>a:a:w::I:U

Figure··3. . Infomlal subsurface stratigraphic units of the Bell Canyon FOmlation, the uppemlostfOmlation of the Delaware Mountain Group.

-134-

R28E R29E R30E R31E R32E R33E R34E R35E

T255

T265

T225

T235 T2350 6 ml

I IP 90 10 km

T245 T245

T225

T265

T255

Figure 4. Lateral distribution and thicknesses of channel sands in the informal Ramsey memberof the Bell Canyon Formation in the northern Delaware Basin (from Lappin, 1988).

Transmissivity values at Cabin Baby-l andDOE-2 ranged from 5 x 10-5 to 6 x 10- 1

ft 2 /day (Beauheim and others, 1983;Beauheim, 1986).

Mercer (1983) revised the equivalentfreshwater potentiometric surface map of Hiss(1976) to include data from WIPP boreholes(Figure 5). The potentiometric contoursgenerally follow the regional structural trend.Flow directions are generally northeastwardacross the WIPP site area (Figure 5).Recharge occurs along the western margin ofthe Delaware Basin. The salinity ofgroundwater within the Bell Canyon increasesfrom fresh to very saline (300,000 mg/L TDS)from the recharge area to the structural troughof the basin (Grauten, 1965; McNeal, 1965),indicating abundant rock-water interaction.The Bell Canyon may discharge into theCapitan aquifer along the northeastern rim ofthe basin, but little is known about Capitanhydraulic heads at those depths.

Castile FormationThe Castile Formation directly underlies theSalado Formation (Figure 1). It is an aquitardthat forms a hydrologic barrier between theBell Canyon Formation and the SaladoFormation.

General Hydrology

Near the WIPP site, the Castile Formation isapproximately 1,200 ft (400 m) (Lappin,1988), and it consists of thick units oflaminated anhydrite/carbonate and halite(Figure 6). The Castile has low permeabilitiesand usually only minute quantities of in situfluid. Regional hydraulic gradients have neverbeen established within the Castile across theDelaware Basin, although localized zones ofbrine and gas under high pressure have beenencountered by several drillholes intersectingdeformed and fractured anhydrite within theCastile. The hydraulic properties ofundisturbed Castile anhydrite and halite havenot been determined because the transmissivity

-135-

Contour Interval Is 100 feet.

10 5 0 10 MILESr-.----I---I--10 5 0 10 KILOMETERS

SCALE

EXPLANATION

\SSSS\ GUADALUPIAN REEF COMPLEX

Figure 5·~ Contour map of the estimated freshwater head elevation for the hydrologic unit in theupper part of the Bell Canyon Formation (from Mercer, 1983).

-136-

Anhydrite III

Z Halite II0-I-<t~a:0 Anhydrite IIu.W...J-I-en<tU Halite I

Anhydrite I

Figure 6. Informal stratigraphic units of the Castile Fonnation near the WIPP site.

is too low to measure with current testmethods. Hydraulic properties of fracturedanhydrite brine reservoirs have been estimatedduring testing of WIPP boreholes ERDA-6and WIPP-12 (Popielak and others, 1983).

Castile Brine Occurrences

Pressurized brine and gas occur in someCastile anhydrite that is fractured anddeformed (Anderson and Powers, 1978).Borns and others (1983) and Borns andShaffer (1985) examined defonnation in theCastile near the WIPP site and in larger areasof the basin. Borns and others (1983)concluded that the Castile structures were dueto gravity-driven salt flowage/defonnation.

Isolated zones of pressurized brine have beenencountered in the Castile at severalhydrocarbon exploration holes and two WIPPdrillholes, ERDA-6 and WIPP-12 (Figure 7)

(Popielak and others, 1983). The brine iscapable of discharging at the surface.Popielak and others (1983) concluded that thebrines originated from ancient seawater withno fluid contributions from present meteoricwaters, based upon analysis of major andminor element concentrations in the brines.The gas and brine chemistries, and someisotopic compositions, are distinctly differentfor each reservoir, indicating isolation and adistinct origin for each reservoir. The brinesare saturated, or nearly so, with respect tohalite. The reservoirs occur in fracturedanhydrites defonned by salt anticlines (Figure8). Most of the brine is believed to be storedin low-permeability microfractures with onlyabout 5 percent of the total brine in largefractures. The volumes of the ERDA-6 andWIPP-12 brine reservoirs are estimated to be3.5 x 106 and 9.6 x 107 ft 3 , respectively.Hydraulic heads within the reservoirs arehigher than in other regional groundwater

-137-

I~

W00

I

T23S

R30E R 31 E R 32 E

EDGE OF CAPITANLIMESTONE

R 33 E

-414•-----500-

•-~

.-609

LEGEND:

• BRINE OCCURRENCEIN CASTILE FORMATION

• BOREHOLE PENETRATINGCASTILE FORMATION

T WHICH DID NOT21 INTERCEPT BRINES

.48 ELEVATION OF TOP OFHALITE II (FEET ABOVE /BELOW MSL)

(II) NUMBER REFERS TOBRINE OCCURRENCETABLE H.I.

NOTES:

I. CONTOURS DRAWN ONTOP OF HALITE IIMEMBER OF CASTILEFORMATION. DASHED WHEREINFERRED.

2. ALL BOREHOLES SHOWNWERE USED AS CONTOURCONTROL POINTS EXCEPTTHOSE WITH UNRELEASEDLOGS.

3. AREA WITHIN HEAVYDASHED LINE SHOWN INFIGURE G-12 IN GREATERDETAIL USING 3EISMICTIME ISOCHRONS.

SCALE

i

o Z MILES

Figure 7. Pressurized brine occurrences in the Castile Fonnation relative to structure on the top of Halite 2 unit (from Popielak andothers, 1983).

DETAIL A-A

ANHYDRITE

HALITE

8. REGIONAL TILTING

\'--­

C. PRESENT CONDITIONS

All

1000

zo~

~ 0~

..J

~ 500~

>~

c:s:..J~

a::

500~~

o

,~

V-)

\0I

DETAIL A-AN.T.S.

Figure 8. Schematic of brine reservoir formation (from Popielak and others, 1983).

systems, and they have been maintained for atleast a million years (Barr and others, 1978).

Time-domain electromagnetic (TDEM)methods suggest that a portion of a brinereservoir underlies the northern andnortheastern most waste-emplacement panels(Figure 9; Earth Technology, 1988).

Salado FormationThe Salado Fonnation is infonnally dividedinto three members: unnamed lower and uppermembers, and a middle member, locallydesignated the McNutt potash zone (Figure10). Important sulfate beds are numbered asmarker beds. About 85 to 90 percent of theSalado is halite (Jones and others, 1973).Unlike the Castile, the Salado extends wellbeyond the Capitan reef onto the NorthwesternShelf and the Central Basin Platfonn. TheSalado has been erosionally removed anddissolved by circulating groundwaters alongthe west and south sides of the DelawareBasin, leaving brecciated insoluble material.

General Hydrology

The Salado is a regional aquiclude; withinmost of the Delaware Basin, the Salado acts asa regional barrier to fluid flow (Lappin, 1988).Regional hydraulic gradients within Saladohalites and interbeds, if present, have neverbeen established. Pressurized brine flows ofsignificant size have never been encounteredduring the hydrologic testing of the Salado(Lappin, 1988), although a slow build-up ofpressure has been observed during thehydrologic testing of some boreholes (Mercer,1987). Mercer (1987) reports a maximumobserved pressure of 472 psi at the WIPP-12wellhead and suggests that the pressures aresupported by a very small volume of brineproduced from anhydrite and/or clayinterbeds.

Pressurized gas has been encountered withinthe Salado in surface boreholes, local potashmines, and WIPP excavations (Lappin, 1988).The origin of this gas has not been firmlyestablished. Mercer (1987) reports a large gasflow (predo~inantly nitrogen) at AEC-8, asurface ··drillhole, of 2 x 107 ft3 with little

apparent depletion of the gas reservoir. Thisinflow occurred roughly one year after originaldrilling. Deal and Case (1987) report that gasexsolves from Salado brines collected in theWIPP underground. The other gasoccurrences may be attributable to exsolutionof dissolved gases during fluid-pressure reliefaround boreholes and excavations.

Where halite-unsaturated groundwaters haveactively circulated through the Salado,dissolution and collapse have resulted (Holtand others, in prep.). On the basis ofgeophysical logs, Holt and others (in prep.)evaluated the extent of Salado dissolution inthe vicinity of the WIPP site (Figure 11).Although Salado dissolution is currently activein Nash Draw and areas adjacent to the PecosRiver, most of the dissolution of Saladomaterials east of the Pecos River developedduring the Cenozoic in response to hydraulicgradients associated with the ancestral PecosRiver (Holt and others, in prep.). The Saladowas dissolved in response to circulatinggroundwater related to near-surfacegroundwater regimes, not by regionalconfined groundwater systems (Holt andothers, in prep.).

Near-Facility Hydrology

The Salado Formation is respondinghydrologically to pressure gradients aroundthe WIPP excavations. Brine inflow to driftsand shafts has been well documented (e.g.,Deal and Case, 1987; Holt and Powers, 1986,in prep; Nowak and others, 1988). Brine hasbeen observed accumulating in boreholes, asweeps on the surfaces of excavations, andseeping from anhydrite and polyhaliteinterbeds in the air-intake shaft (Deal andCase, 1987; Holt and Powers, 1990). In mostboreholes, fluid-inflow rates vary from a fewhundredths to a few tenths of a liter per day(Deal and Case, 1987; Nowak and others,1988). Brine inflow into boreholes usually isminimal following the initial drilling, increasesto a maximum, and then slowly decreases withtime (Deal and Case, 1987). Holes drilleddownward generally accumulate more brinethan holes drilled upward. Brine inflow varieswith stratigraphy (Deal and Case, 1987).

Bredehoeft (1988) suggested that the Saladoshould be considered a saturated medium.Stratigraphic and lateral variability within the

-140-

500 1000 FEET

'00 ZOO 300 METERS

C/'t

$

Section Corner

Drill Hole

LEGEND

Storage Pond

Salt ana TOPsoil Storage Area

Transmitter Loop Corner

Areas Where ConauctorIS at Castile Fm aeptns

Transmitter Loop Center(500 m x 500 m Loopsl

Control POint«(

o

•

300 ,,)0 100

++~D~

1000 500

o

o

o.,373

r0~~,~~r.-~,~~1

l' _'-..11'-:" / I' ,. .... 9 ... .1" f '~I:IIUI ".'313

IIIIII--------

;:/~;?:~~:~~",_:,(; 0,'.'- '~'~ ./: ~,J I, _~~:" I ::

["".~;-ptA·1iT-SiTit}'·REA :i~i-" ::I c- J..-.:.J:.:>.' II

: C&SH ~ :-: :;':c?\~\ ::

~ 0 Shaft t:' ~ . :<1.' \~'. ::

"

"

"I

I,j

""'."

,; 0 _~ 0 J,f=======~,_'============ I

I ,,'/ "

J

I

I

II/

TSON - u

SOON - 0

250N - 0

OOON- 0

I......+::-......

I

o

Figure 9. Contour map to depth of conductor from time domain electromagnetic (IDEM) sUlVeys of the WIPP site, indicating possiblebrine locations (from Earth Technology Corporation, 1988).

RUSTLER FORMATION

wzaN

a: M8103wQ..

MB109Q..:::>

VACA TRISTE SS MEMBER....

W ~0....

Z I-

a z ~

N UJ e"I- :::)

I Cf) 0(j) ~UJ a:

~e"z ~=,0 I-

a UNION ANHYDRITE ONa:UJ M

Q.. MB123/124 ~

Z ~a: ......... 1-00 ..... co

I-~- :::> Cf).- z c:::

« 0 -~

u..~0:0 MB126U.

0Qc(--J a:« wen co

~w~

COWDEN ANHYDRITE MEMBERa:w~ I NFRACOWDEN SALTa..J

~--LA HUERTA SLTST.

FLE TCHER ANHYDRITE MEMBER

CARLSBA~LIMESTONE CASTILE FOR MATION

Figure 10. General Salado Fonnation stratigraphy in the vicinity of the WIPP site (from Holt andothers, in prep.).

-142-

R32ER33E

H-l0•

o .u 0u>- -eo 1&.1

~l...l

NO SALADODISSOLUTION

H-12•

-H-l1

P-17 H-17• •

R31E

R32E

DOE-2•

WIPP sITeBOUNDARY

WIPP-13 H-5•

• WIPP-12H-lI! WIPP-18

• ·WIPP-19WIPP-22IwIPf.-21H-2· ·H-l H-15

H-3· OOE-l P-18eH-14· •

H-8-

R31ER30E

ZONE

'" OF\ SA~ADO H-7

DISSOLUTION-

\ /........ _--.,.,.,.".

,.'ERCE CANYON"

e

I/

//

o 3 Mil S.'-----'

lAGUNA ORANDEDE lA SAL

T21S

T24S

T22S

Figure 11. Interpreted extent of Salado dissolution (from Holt and others, in prep.).

salt is high and currently precludes any directinterpretation of the degree of saturation in theSalado.

Continuing hydraulic testing shows that thehydraulic conductivity of argillaceous haliteranges from about 4 x 10-9 to 7 x 10-8 ft/day(Beauheim and others, in prep.). Thehydraulic 99nductivity of relatively pure halitehas p~oven to be immeasurably low «10- 10

ft/day), if it exists at all. Marker Bed 139, ananhydrite interbed approximately 3 ft thick,has hydraulic conductivities ranging from 1 x10-7 to 2 x 10-6 ft/day, apparently dependingon the amount of fracturing present. Porepressures show a distinct gradient towards theunderground facility, with pressures as highas 1700 psi being measured at distances of 3Sft and beyond from the excavations.

-143-

Generally, penneability increases aro.und thefacility within 5-10 ft due to fractunng andpossibly matrix dilation (Borns and Stormont,1988). Fractures develop parallel to t~e

excavation, while low-angle fractures form Inthe corners of the rooms (Westinghouse,1988). Some high-angle fractures develop inMarker Bed 139, and pre-existing fracturesmay open further (Westinghouse,. 19~8). Thefracture development process IS time andexcavation-geometry dependent (Borns andStormont, 1988).

The brines which flow into the WIPP facilityare not derived from fluid inclusions, butinstead are grain-boundary fll;li~s withresidence times of at least several milhon years(Stein and Krumhansl, 1986). Geophysicalmeasurements suggest the far-field b:inecontent of halite to be around two weIghtpercent and a near-field brine conten! ofapproximately one weight percent (PfeIfer,1987; Hudson, 1987).

Rustler FormationThe Rustler Formation is the uppermost ofthree Ochoan evaporite-bearing formations inthe Delaware Basin. The Rustler has beenextensively characterized si~ce 1972 as .itcontains the most transmissIve hydrologIcunits overlying the WIPP facility horizon (assummarized in Powers and others, 1978;Lappin, 1988). The Rustler has a variablelithology consisting of inter1?edded .su~fates,

carbonates, clastics, and hahte. WithIn theRustler, five hydrologic units have beenidentified: the unnamed lower member/Saladocontact area, the Culebra Dolomite Member,the Tamarisk Member claystone (M-3 of Holtand Powers, 1988), the Magenta DolomiteMember, and the Forty-niner Memberclaystone (M-4 of Holt and P~wers, 1988)(Figure 12). The Cule?ra IS the ,mosttransmissive and regionally Important unit andhas been the focus of Rustler hydrologicstudies.

The Rustler varies in thickness from tens offeet, where exposed and subjected to solutionand erosion, to 560 ft, in the northeastern partof the Delaware Basin (Holt and Powers,1988; P~wers and Holt, this volume). Wherethe Rustler crops out or is in the shallow

subsurface, it has been extensively altered bynear-surface groundwater and karst processes(Holt and others, in prep.). Karst process.esdo not affect the Rustler across the WIPP sHearea, although evidence of Rustl~r karst, isabundant in Nash Draw and along its margIns(e.g. WIPP-33, about 1 mile east ofLivingston Ridge) (Bachman, 1987; Holt andothers, in prep.).

Unnamed Lower MemberHydrology

The unnamed lower member of the Rustlerconsists of interbedded siltstone, sandstone,halite and anhydrite (Figure 12) (Holt andPowe;s, 1988). Holt and others (in prep.)recognize two confining beds and two water­producing units within the unnamed lowermember.

Holt and others (in prep.) considered thesulfate unit at the base of the Rustler to be thelowermost confining bed within the Rustler,although this hydrologic role is mostly due tolow permeability halite in the Salado. Whe~e

dissolution has affected the upper Salado, thISstratigraphic unit is hyd:~logically

compromised, and the lower conf~nIng bed forwater-bearing unit 1 is the top of Intact Saladohalite.

Water-producing unit 1 (WPU-l) o~ Holt andothers (in prep.) consists of the bIoturbatedclastic interval and the transitional sandstone,and it includes the M-l/H-l interval whenhalite is not present. WPU-1 is .co.nfined byanhydrite. The bioturbated clast1~ u~tervaiis

the principal water-bearing zone.w~thIn.WPU­

1. The siltstone and sandstone WIthIn thIS zoneare argillaceous and cemented with halite (Holtand Powers, 1988; Holt and others, in prep.).The transitional sandstone is also cementedwith halite and locally with sulfate (Holt andPowers, 1988). The M-l/H-l i.nter~al

contains bedded and argillaceous halIte wHhsome halitic clastic rocks. Halite cements aredissolved in the units to create secondaryintergranular porosity and permeability.Dissolution of halite from the upper part of theSalado causes structural collapse of, andupward stoping through, overlyin? units.Fracture porosity develops around dIsplacedblocks of collapsed material, and the th~ckness

of the water-bearing zone increases to Include

-144-

FORMALGEOLOGIC

NOMENCLATURE

DIVISIONS OFHOLT AND POWERS

(1988)

HYDRO­STRATIGRAPHIC

DIVISIONS

DEWEY LAKE RED BEDS

zot­<{

~a:ou..a:w.....Jt­(j)::)a:

FORTY-NINER MEMBER

MAGENTA DOLOMITEMEMBER

TAMARISK MEMBER

CULEBRA DOLOMITEMEMBER

UNNAMEDLOWER MEMBER

SALADO FORMATION

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

- - _._.

SIL TSTONE/CLAYSTONE WATER-PRODUCING

ANHYDRITE 5CONFINING BED 6

(A-5)

MUDSTONE/HALITE 4 WATER-PRODUCING UNIT 5IM-4/H-4)

ANHYDRITE 4CONFINING BED 5

(A-4l

MAGENTA DOLOMITE WATER-PRODUCING UNIT 4

ANHYDRITE 3(A-3l CONFINING BED 4

MUDSTONE/HALITE 3WATER-PRODUCING UNIT 3(M-3/H-3)

ANHYDRITE 2 CONFINING BED 3(A-2)

CULEBRA DOLOMITE WATER-PRODUCING

MUDSTONE/HALITE 2 UNIT 2

(M-2/H-2)ANHYD~ITE 1 CONFINING BED 2IA-l ------------

MUDSTONE/HALITE 1CONFINING WHERE H-1

(M-l/H-1lWATER-PRODUCING

WHERE M-l

TRANSITIONAL SANDSTONE------------

ZOt-lL

BIOTURBATEDWATER-PRODUCING

CLASTIC INTERVALUNIT 1

(SILTSTONE)

SULFATE UNIT CONFINING BED 1===HALITE CONFINING BED

Figure 12. Relationship of lithostratigraphic and hydrostratigraphic units for the RustlerFormation (from Holt and others, in prep.).

the interval hydraulically connected byfractures. After collapse, the evaporitecements within the collapsed blocks are morereadily dissolved. Hydrologic testing ofWPU-l is difficult as the borehole may onlyeffectively intersect a few fractures and theperm~ability between collapse blocks may belimited by clay skins developed on the marginsof blocks during the collapse process.

Transmissivity values for WPU-1 (Mercer,1983; Beauheim, 1987) generally show atransmissivity increase in the direction of NashDraw (Figure 13). Holt and others (in prep.)suggest that the transmissivity values fromareas where Salado dissolution and collapse ofoverlying strata have occurred may not berepresentative of the entire WPU-1 intervalthickness at the tested locations.

-145-

WIPP-28. 087 e

..~.../~

"11/"~~.

"/~

'''{

=~~,

.,~

.~~

\~...~

e WIPP-27. 2 x 10-'

_eWIPP-30. 0 2

~..

....~.".s

.~,~

e P-1S. 3 x 10- 5

~WIPP-SITE

BOUNDARY

.....P-17. 2 x 10-'

H-4c. 6 ~ 10-'

H-16. 2 x 10-·.......

H-2c, 1 x 1 ee _eH-l. 3 x 10-'

H-3~3 x 10-'

• P-15. 4 x 10-'

H-6c. 3 x 1O-3r -,~,. H-5c. 3 x 10- 5

"$"WIPP-25. 5 e ~,-

~

t-~ ~,.'

.w,. ~".,

WIPP-26.04'*''/11 ',1.

'fl" '~n", ,'?;"';:

4.~,

~'~,

.}~"

~.""N

e WIPP-29. 8

H-7c.073e

H-l0c, 9 X 1O- 5 e

~­~.

"!'!..l\.. ~~.

~'f'-

,"-?... ,

9'i';,

!i

H-9c_ 2 x lO-"e

eH-Sc. 3 X 10 3

LEGENDi-----WELL TESTED

.P-14 5 x 10-2 TRANSMISSIVITY OF THE UNNAMED LOWER. .. MEMBER AND/OR RUSTLER-SALADO

CONTACT RESIDUUM, ft 2/day

aI

2 3

SCALE

4 miI

Figure 13: . Transmissivity of the unnamed lower member of the Rustler and/or residuum along theRustler/Salado contact (from Holt and others, in prep.).

-146-

Few measurements have been made of thestabilized water level or fluid pressure ofWPU-l (Figure 14). Although freshwaterhead calculations are somewhat limited due tothe quality of specific gravity measurements,estimated freshwater head values are probablyaccurate to within 10 ft (Holt and others, inprep.). The freshwater heads in those wellscompleted within the collapse materialassociated with Salado dissolution do notappear to differ significantly from thehydraulic heads in wells completed in intactRustler siltstone, indicating good lateralhydraulic connection. The heads areconsistent with Mercer's (1983) interpretationthat flow through the low transmissivitysections of WPU-l is generally westerly orsouthwesterly across the WIPP site towardsNash Draw, while flow within Nash Draw isgenerally southwesterly towards Malaga Bendon the Pecos River.

Holt and others (in prep.) divided the area ofWPU-1 into three distinct hydrogeologiczones (Figure 15). The first zone ischaracterized by low permeabilities, no Saladodissolution, and intergranular porosity andpermeability. Hydraulic properties in zone 1are related to the dissolution of halite cements,and permeability generally decreases towardthe east across the WIPP site area. WithinZone 2, WPU-1 is characterized bypermeabilities higher than in Zone 1, Saladodissolution with varying amounts of collapse,and intergranular porosity coupled withfracture porosity proportional to the amount ofcollapse. Partial confining conditions existwithin zone 2, and the zone is not actively tiedto the surface groundwater regime, as varyingamounts of overburden isolate WPU-l. Zone3 is actively recharged by the surfacegroundwater system and is confined toportions of Nash Draw. Overburden islimited, and the zone may be partly confined.Dissolution and collapse are extensive withinzone 3, and evaporite cements within collapseblocks are mostly dissolved. It is expectedthat the hydraulic properties of WPU-l havegreat lateral and vertical variability due to thecomplex nature of collapse and brecciation.Part of zone 2 and all of zone 3 can beconsidered within Lang's (1938) brineaquifer. Modern flow within zones 1 and 2 isgenerally west toward Nash Draw, while flowin zone 3 is southwest along the axis of NashDraw toward Malaga Bend.

Confining bed 2 consists of the transitionalsandstone, where evaporite cements are notdissolved; the M-l/H-l interval, where haliteis present as cements and interbeds; andanhydrite 1 (Holt and others, in prep.). Thisconfining unit is most effective across theWIPP site area where no dissolution hasaffected the Salado (Figure 11; Holt andothers, in prep.). Anhydrite 1 is halitic acrossmuch of the WIPP area and displays somefractures (Holt and Powers, 1988). WithinNash Draw and other areas of extensivedissolution of Salado evaporites, dissolutionof the halite within A-I and additionalfracturing due to collapse may hydraulicallybreach A-I, allowing vertical leakage betweenwater-producing units 1 and 2.

Mudstonelhalite 2 (M-2/H-2) and the CulebraDolomite Member constitute water-producingunit 2 across the WIPP site area (Holt andothers, in prep.). East of the WIPP site, M­2/H-2 contains halite and is considered to bepart of confining bed 2. M-2/H-2 consists ofinterbedded sulfatic mudstone and siltstoneacross and west of the WIPP area (Holt andPowers, 1988). It is probably in directhydraulic communication with the Culebra andacts as a source of leakage to the Culebraduring hydrologic tests where the hydraulichead of the Culebra is lowered (Holt andothers, in prep.). M-2/H-2 is discussedfurther with Culebra hydrology.

Culebra Dolomite Member

The Culebra is the most transmissivehydrologic unit overlying the WIPP facilityhorizon. It is considered to be the mostimportant groundwater-transport pathway forradionuclides that may escape from the WIPPrepository to the accessible environment.Near the WIPP, the Culebra is finelycrystalline, locally argillaceous andarenaceous, vuggy dolomite roughly 25 ftthick (Holt and Powers, 1988). The Culebrahas been the subject of extensive hydrologictesting, summarized by Mercer (1983) andLappin (1988). Culebra groundwatergeochemistry has been studied in detail (e.g.,Siegel and others, 1990), and the Culebrageology has been extensively characterized(e.g., Holt and Powers, 1988; Holt andothers, in prep.).

-147-

I

N

OP-18

--WIPP-SITEBOUNDARY

OOOE-l

OWIPP-13

oWIPP-12H-180 oWIPP-18

H-16.~ 8~:~~:;~>3130- gWIPP-21

>3035w 0" OH-15H-2c. >Ji3i H-l EROA-9

OH-3OH-14

Q

OWIPP-27./~.

'/~

..,~·~/1

~'

.-:£ ~~~. 0~ I­~

,,\.~ co~ 2937 oOOE-2-.?- ~', H-6c. ~300::::'2~II'---W---...:L.::"::':;"~-H-_-5""0

WIPP-25~ _ ~ 0. >3035 ~ _'\..

~\' "~\\. ~ "-

,~h,~ ~

b.. ~t, P-14. ~wl . WIPP-26~~" 3037

. :> 2964 "'~/"

'ljll' •~

,~

"'~~

,\\\~

,'~~~

A..~

~Ilil' "\\~'

oCAB'N BABY

P-17.~. OH-17>3020

2959.WIPP-29. ­

2975

H-7e.~.2973

P-15 OH-ll

OH-12

~\\, .~.~'~~l .~\""

~~\....;;-;;.H-l00

oH-9

oENGLE

.H-8C.~3013

LEGEND

r-r=WELL TESTED

.r--MEASURED WATER-LEVEL, ELEVATION. ft amsl

.WIPP-28 3055. 3097L-ESTIMATED FRESHWATER-HEAD

ELEVATION. ft amsl

2 3! !

SCALE

4 miI

o OTHER OBSERVATION WELLS

Figure 14. Measured water level and estimated freshwater head elevation of the unnamed lowermember of the Rustler and/or residuum along the Rustler/Salado contact (from Holtand others, in prep.).

-148-

H-5

T21S

T22S

T24S

//

//

-lAGUNA 01lANDE

Of lA SAL

PIERCE CANYON"

o 3 MIL S.'---_....'

R31E

R30E

ZONE 2

H-8-

R32ER31E

DOE-2I -8 •

WIPP-13•.WIPP-12

H-lIl WIPP-18- ·WIPP-19

WIPP-22IwIPf-21H-2- -H-1 H-15

H-3- DOE-1-H-t.. •

-H-11

P-17 H-17- -H-12-

P-18•

R33ERJ2E

o 0o 0> •o ua~I..l

ZONE 1I

NO SALADO

DISSOLUTION

H-tO-

Figure 15. Extent of hydrogeologic zones within water-producing unit 1 (WPU-1) of the RustlerFormation (from Holt and others, in prep.).

-149-

Culebra Hydrology

The Culebra is the hydraulically mostsignificant portion of WPU-2. M-2/H-2probably provides some leakage into theCulebra during pumping tests and is thereforeconsidered part of WPU-2. WPU-2 isconfined by anhydrites in the unnamed lowermember and the Tamarisk Member. WithinNash Draw and adjacent to Nash Draw,collapse following evaporite dissolution mayhave decreased the effectiveness of theanhydrites as confining beds.

Water levels in 60 wells completed to theCulebra at 41 drilling-pad locations have beenmeasured on a regular basis (Holt and others,in prep.). Culebra water levels at and aroundthe WIPP site have been affected, however,by continuous drainage into one or moreWIPP shafts (since late 1981), as well as bynumerous pumping tests and water-qualitysampling exercises conducted since 1980(Haug and others, 1987).

Cauffman and others (1990) thoroughlyreviewed Culebra water-level data, borehole­fluid density data, and WIPP-related hydraulicstresses, and estimated the undisturbedfreshwater heads at 35 wells (Figure 16).Flow is generally to the south (Crawley, 1988;LaVenue and others, 1988) across the WIPPsite area (Figure 16). Davies (1989) andCrawley (1988) suggest that flow directionssouth of the WIPP site may have a largereasterly, down-dip component than ispredicted considering only freshwater heads.

The transmissivity of the Culebra across theWIPP region varies by six orders ofmagnitude (Figure 17) due to varying degreesof fracturing within the Culebra. In thevicinity of the WIPP, the Culebra has beentested in 41 locations, including the combinedtesting of the WIPP project and ProjectGnome (Beauheim, 1988). The highesttransmissivity value reported for the Culebra is1.25 x 103 ft2/day at WIPP-26 within NashDraw. The lowest value reported is <0.004ft 2 /day at P-18. Where Culebratransmissivities are less than 1 ft2/day, theCulebra behaves as a single-porosity mediumduring pumping and slug tests. Wheretransmissivities'are between 1 and at least 100ft2/day, the Culebra behaves as a double-

porosity medium, with matrix and fractureporosity sets. Where the Culebratransmissivity exceeds 100 ft2/day, double­porosity behavior is less apparent.

LaVenue and others (1990) calculated Culebratransmissivities across the WIPP site areabased upon steady-state calibration against anestimate of the preshaft distribution offreshwater equivalent heads (Figure 17).Aside from the apparent increase oftransmissivity toward Nash Draw, severalother relationships are important.Transmissivities are relatively high in a wideregion south of the WIPP site and west inNash Draw. A high-transmissivity "finger"penetrates the southern border of the WIPPsite. Finally, a zone of higher transmissivityoccurs north of, but does not penetrate, theWIPP site area.

Culebra Groundwater Chemistry

Culebra groundwater geochemistry isextremely variable across the WIPP siteregion. The total dissolved solids (TDS)within the Culebra groundwater increase fromless than 5,000 mg/1 southwest of the WIPPsite to greater than 200,000 mg/1 northeast ofthe WIPP site (Figure 18) (Siegel and others,1990).

Culebra water has been typed by Siegel andothers (1990) and Holt and others (in prep.)on the basis of major element concentrations(Figure 19). Using different sample sets andslightly different criteria, both Siegel andothers (1990) and Holt and others (in prep.)report four coinciding "type areas" of Culebrawater (Figure 19). The easternmost type areaof Siegel and others (1990) and Holt andothers (in prep.) (Zone A and Type Area 1,respectively) is characterized by highly salineNaCI brines rich in Mg and Ca. Culebra waterfrom Zone B of Siegel and others (1990) andType Area 2 of Holt and others (in prep.) isrelatively fresh, with calcium and sulfate as thedominant solutes. NaCI-dominated waters ofvariable compositions occupy Zone C andType Area 3, with the salinity of this waterincreasing to the east (Siegel and others, 1990;Holt and others, in prep.). Holt and others (inprep.) suggest that mixing of the highly salineNaCI brines in Type Area 2 with the sulfate­rich water in Type Area 1 may have generatedthe Type Area 3 water. Located in the western

-150-

r: ,,'

I (, I~ !

\!:J)

I 3 MILES

SCALE

:i

',1

//

/

U'l(

I

( ~',1

-' ))

I

EX PLANATION

-'-- -~~-I' .

, I, i !

III

;1-'

-3000- FRESHWATER-HEAD ELEVATION CONTOUR,CONTOUR INTERVAL = 25 FEETFROM LA VENUE, ET AL., 1988

• WELLS TESTED

GENERAL DIRECTION OF GROUND-WATER FLOW~ FROM CRAWLEY, 1988-. FROM LAPPIN, 1988

Figure 16. Estimated Culebra Dolomite Member freshwater heads and flow directions (from Holtand others, in prep.).

-151-

R3tE R32E R33ER30E R31E

0 R32E0U U

~ ~

T/f;.I~

21s •

00

I--~

'",.~

<:).

Q~

~~T22 ~,.S

I/

//

•LAGUNA ORANDE

DE LA SAL

T23S

t-iT

2.PIERCE CANYONS •

, 0 3 MIL S, ,

Figure 17. Calculated transmissivities for the Culebra Dolomite Member (ft2/day) (from LaVenueand others, 1988).

-152-

, ):

/

..) 11 (i

(~1!

I 3 MILES

SCALE

I,/

[ 11

.. '

\I

"',.\ '

~f:j(;),

\(I.

\/1. 0 /,,1" I

. /)\ )III

t' - ). I " ' .. ~ ., ~,

•

•

EXPLANATION

, \,1

• WELLS TESTED -10,000- EOUAL CONCENTRATION CONTOUR LINE, mg/l

Figure 18. Distribution of total dissolved solids (IDS) in groundwater of the Culebra DolomiteMember (from Holt and others, in prep.).

-153-

..

J' ,"

SCALE

•

"t :I

, i,.'j

I '

•

•I

( ONE ~)*• TYP AREA 1

• •• ;DS> 100,000'mgll

NaCI TYPE WATER

•

\,I

(ZONE C)I*

TYPE AREA 3

(

(ZONE B)*TYPE AREA 2

\

. /

[ll

(

\.

')

'I(

\

)I

I

\\

3 MILES

'. \ TDS<10,000 mgll

CaS04-TYPE WATER.

.;

I

(ZONE D)* i

TYPE AREA 4 ICONTAMINATED BY; -"

POTASH MININGEFFLUENT

I ~WI)

IEXPLANATION

• WELLS TESTED

Figure 19'. HyCirochemical type areas for the Culebra Dolomite Members (from Siegel and others,1990, and Holt and others, in prep.).

-154-

half of Nash Draw, anomalously highsalinities and potassium contents characterizeZone D and Type Area 4 groundwater (Siegeland others, 1990; Holt and others, in prep.).The water quality in Zone D and Type Area 4is dominated by effluent from potash refiningoperations (Siegel and others, 1990; Holt andothers, in prep.).

Ramey (1985) and Siegel and others (1990)noted that modern flow directions within theCulebra do not appear consistent with themodern salinity distribution. The IDS withinthe Culebra generally decreases in the directionof flow, which is not a typical steady stateIDS/groundwater-flow relationship. Isotopicevidence (Lambert and Harvey, 1987; Siegeland others, 1990), as well as modem Rustlertransmissivities and head distributions,preclude using modern vertical recharge to theCulebra (Lappin, 1988), to account for themodern flow/salinity relationship. Onepossible mechanism involves fluid flow to thenortheast during past recharge followed bymodern system draining and discharge to thesouth (Lambert and Harvey, 1987; Holt andothers, in prep.). Lambert and Harvey (1987)interpret the hydrologic setting for the WIPParea as transient, with no active recharge to theCulebra, rather than as steady state.

Saturation indices for several evaporiteminerals in Culebra groundwater werecalculated by Holt and others (in prep.) and bySiegel and others (1990). The saturationindex of a particular mineral equals zero atsaturation, while positive or negative valuesindicate that the groundwater is abovesaturation or below saturation, respectively.Holt and others (in prep.) found Culebragroundwater to be undersaturated with respectto gypsum, although the degree ofundersaturation varies (Figure 20). Valuesbetween +0.1 and -0.1 can be consideredsaturated. The Culebra waters areprogressively more undersaturated withrespect to gypsum from east to west, and azone of less saturated waters penetrates theWIPP area from the south. The overall trend issimilar to the transmissivity distribution in theCulebra, with the degree of undersaturationincreasing with transmissivity. However,Siegel and others (1990), using a differentgeochemical model, found no undersaturationof Culebra -waters with respect to gypsum.

The discrepancy between the two modelsremains to be resolved.

Culebra Hydrogeology

Open and sulfate-filled vugs and fractures arelocally abundant in the Culebra (Holt andPowers, 1988). The large vugs are principallyconcentrated in the lower portion of theCulebra (Holt and Powers, 1990). Most ofthe penneability is along fractures, andtransmissivity appears to be related to degreeof fracturing. Holt and others (in prep.)evaluated the hydrogeology of the Culebra andexamined the Culebra for depositional andpost-depositional controls on its hydrologiccharacter. The following is a summary oftheir efforts.

The Culebra was deposited in an areallyextensive carbonate-rich lagoonal environmentfollowing a rapid marine transgression over asalt pan/saline mudflat environment (Holt andPowers, 1988). As the water-depth wasstable over most of the WIPP area, theCulebra was unifonnly deposited with slightincreases in thickness in those areas wheredissolution of the underlying saltpan/salinemudflat evaporites had occurred (Holt andPowers, 1988). The lagoonal environmentwas persistent for all of Culebra depositionwith only episodic shoaling contributing minoramounts of lateral and vertical variability, sofacies changes do not affect the hydrologicproperties of the Culebra. Slumping ofCulebra sediments occurred in response tocompaction of algal laminae, dewatering, anddissolution within and compaction of theunderlying sediments. This deformationprovided a vertical to subvertical fabric forlater fracture reactivation.

Much of the Culebra sediment was uniformlyfine-grained carbonate mud, lithified toproduce finely crystalline dolomite of lowpermeability. Poorly sorted, commonly veryfine-grained, dolomitic and possibly organic­rich interbedded sediment was locallyinterspersed by bioturbation and sedimentslumping throughout the fine-grainedcarbonate mud. This material provided apathway of higher permeability for circulatingwaters containing solutes for syndepositionaldisplacive growth of anhydrite and gypsumwithin the soft Culebra sediment. Sulfatenodules and crystals grew more often within

-155-

(\\\

,-".1

•

r1'1 II

Pl/,I

r--

-;1'> r'

, '

r: \'J

"'1,1

/,/

•

_.... ~-------,,"" ...../r~--

3 MILES

SCALE

!",' \/_."

, \

I \, '\r L./_\J-" '---

. I M'![)II\It·,I-I/ \.j, ,(--...r "

..) I

I

i/ I'

IiIII

I

/

.~~({;

~~

~~

~,~O' ; ~

IIdI •

\•

•

EX PLANATION

• WELLS TESTED

Figure 20' Contour map of gypsum saturation indices of Culebra waters (from Holt and others, inprep.).

-156-

this poorly sorted material than within theuniformly sized carbonate mud. Most of theinterparticle porosity within this poorly sortedmaterial was filled with sulfate cement.During early diagenesis, much, if not all, ofthe gypsum in the Culebra was replaced byanhydrite. During late-stage diagenesis, thehydraulic properties of the Culebra wereenhanced when diagenetic pore-fillingmaterials dissolved, opening additionalporosity.

Gypsum and anhydrite commonly fill poreswithin the Culebra near the WIPP site. Sulfate(usually gypsum) occurs as cements, passivevoid-fillings, nodules, and incrementalfracture-fillings. A clear pattern of alterationoccurs within the Culebra from east to west.Anhydrite vugs and crystals are replaced bygypsum, and gypsum occurs as incrementalfracture fillings within the Culebra dolomite.Further west, all sulfate consists of gypsum,and gypsum fracture-fillings, cements, andvug-fillings are partly to completely dissolved.Continuing west, gypsum is almost totallyremoved, and many of the vugs and vugcomplexes are collapsed, thereby increasinglocal fracture porosity. In the westernmostpart of the region, no gypsum occurs. Theareal distribution of this relationship is bestshown by the percentage of natural fracturesfilled with gypsum (Figure 21). This mapshows a general inverse relationship betweenthe presence of gypsum pore-fillings andtransmissivity (Figure 17). Presence or

'absence of gypsum pore-fillings may,therefore, be one of the factors controlling thedistribution of transmissivity within theCulebra.

The effective porosity within the Culebra isrelated to fractures, and the transmissivityvalues at various locations indicate degree offracturing. This allows the degree offracturing to be compared with other geologicinformation on a regional scale. The Culebrafractured significantly in response todifferential unloading, dissolution ofevaporites from above or below the Culebra,and dissolution of large vug-fillings or fillingswithin zones of vugs in the Culebra. Regionalor local tectonic activity has not causedsignificant fracturing within the Culebra.

Culebra cores yield little informationconcerning the degree of fracturing due to

variable core recovery and zones of poorrecovery. Culebra descriptions from threeshafts at the WIPP show that the majority ofthe fractures at the WIPP site are subverticaland occur within a very vuggy zone near thebase of the Culebra (Holt and Powers, 1984,1986, 1990). These fractures usually extendfrom vug to vug (Holt and Powers, 1990).This zone can be recognized within Culebracores as a zone of poor or no recovery.Horizontal fractures, parallel to beddingplanes, occur throughout the Culebra in coreand the WIPP shafts (Holt and Powers, 1984,1986, 1990). High-angle subvertical fracturesare commonly intersected in core near theupper part of the Culebra and may be verticallypersistent for several feet. As these nearlyvertical fractures are common, the density ofthese fractures must be relatively high, at leastlocally.

Externally driven changes in the stress fieldwithin and around the Culebra are amplifiedby the different mechanical responses ofunderlying claystone to these changes. Theclaystone below the Culebra is pliable,cohesive, and moist. Abundant arcuate andvariably oriented slickensided surfaces anddeformed strata within the claystone attest to acomplex history of deformation. Theclaystone accomodates changes in stressrelatively rapidly by deforming ductilely andshearing along some surfaces. This rapidresponse to changes in stress locally increasesthe strain and deviatoric stress within theoverlying Culebra dolomite, resulting in brittledeformation of the Culebra.

The Culebra dolomite is much more likely tofracture in response to changes in stress thanare the overlying anhydrites, due toinhomogeneities with the Culebra. Horizontaland subhorizontal bedding planes and near­vertical syndepositional shear planes arebounded by clay-sized material within theCulebra. Fracture response to stress regimevariability will initiate along these pre-existingplanes of weakness rather than in the morehomogeneous dolomite. In addition, abundantlarge sulfate-filled and open vugs in the lowerportion of the Culebra behave differently fromthe dolomite when the Culebra is strained.Fractures propagate from vug to vug, oftenalong syndepositional shear planes, as thestrength of the Culebra is exceeded. Theanhydrites contain fewer inhomogeneities, and

-157-

/

//

•

3 MILES

SCALE

';',.\

\\

\\,

. ~\\

A,I(lil.' \,

>". I

\ .. f

".'".,"".'

11'"

[II'

\

•

(I

'-"'1 "': r-,

" "

:',:, :

I

I·\1~ ~J J\1), r N M 0

I' ".WIPP S TE

•

\ .~---++--+-' I \

\1/

1\

/ I I. \ \~ J ~ 0 ~,

i 1 ~ '\ \

I, '\ \\ \ ". \,

EXPLANATION

• WELLS EXAMINED CONTOUR INTERVAL 10% 5% LINE SHOWN FOR CLARITY

Figure"21. -Percentage of natural fractures in the Culebra Dolomite Member filled with gypsum(from Holt and others, in prep.).

-158-

the strength across bedding planes within theanhydrites is greater than within the dolomite,because clay is absent along the anhydritebedding plane surfaces and microscopicanhydrite crystals interlock across thesesurfaces.

As tectonism is not a feasible externalmechanism for generating major changes inthe stresses acting on the Culebra at the WIPPsite, other external mechanism must controlexternal stress changes. Within the regionalcontext, only differential unloading anddissolution of evaporites from within andbelow the Rustler are capable of effectingexternal stress-field changes. Internal stress­field changes can be generated from within theCulebra by dissolution of pore-fillingmaterials.

Much of the fracturing within the Culebra inthe vicinity of the WIPP site can be attributedto unloading (Holt and others, 1988). Anisopach map of the Culebra overburden showsa general decrease of overburden from east towest, with the overburden north and south ofthe WIPP site slightly less than that across theWIPP site (Figure 22). Culebra transmissivity(Figure 17) is generally higher where there isless overburden. As the transmissivitycontours do not correlate precisely with theamount of overburden present, othermechanisms must contribute to the variationsof transmissivity and the degree of fracturingin the Culebra across the WIPP area.

Aside from unloading, the most significantcontrol on fracturing within the Culebra is thedissolution of evaporites from strata above andbelow the Culebra. The overlying stratacollapse, and stress in the underlying strata isrelieved. Both collapse and stress relief cancause or enhance the fracturing within a brittleunit such as the Culebra. Holt and others (inprep.) evaluate the extent of dissolution fromthe Salado and Rustler on the basis of core,shaft, and geophysical log data (Figure 23).Salado dissolution affects the area south andwest of the WIPP site, while Rustlerdissolution is concentrated along thedepositional margins of halite within theRustler.

The uppermost halite within the Salado isdissolved in the area in and around Nash Drawand in a region south of the WIPP site.

Salado dissolution parallels the course of theancestral Pecos River and its major tributaries,such as Nash Draw. Salado dissolution washydrologically driven by the groundwatersystem associated with the ancestral PecosRiver and its major tributaries. This systemwas most active during the Cenozoic, althoughdissolution of Salado halite continues to occurwithin Nash Draw. Culebra core from areasof Salado dissolution shows evidence offracturing and high transmissivity values dueto collapse.

Salado dissolution close to a halitedepositional margin within the Rustlerincreases the potential for dissolution of halitewithin the Rustler. Collapse associated withSalado dissolution tends to increase thevertical hydraulic connection between Rustlerwater-bearing zones and Rustler intervalscontaining halite. One such zone penetratesthe southern part of the WIPP site. Featuresattributable to dissolution of halite andattendant collapse ~e found within the intervalM-3/H-3 (Holt and Powers, 1988). Thiszone, although not well defined, correspondswith the highly transmissive zone within theCulebra in the southern part of the WIPP site.As dissolution has not occurred below theCulebra at this location, fracturing respondedto stress-relief following dissolution.

Holt and Powers (1988) observed collapsefeatures within the M-3/H-3 interval at WIPP­13 at the northwest comer of the WIPP site.The collapse features at WIPP-13 may be dueto late-stage dissolution of a thin outlying M­3/H-3 halite pan with little or no directconnection to the main M-3/H-3 halite pan(east of the WIPP site). This zone correspondswith high Culebra transmissivity values.

The easternmost zone of potential dissolutionwithin the upper Rustler lies along the limits ofH-3 and H-4, representing the depositionalmargins of halite in the M-3/H-3 and M-4/H-4intervals. East of the zone, halite is present,while west of the zone, halite does not occur.In the vicinity of the WIPP site, halite wasremoved from the M-3/H-3 and M-4/H-4intervals syndepositionally and therefore couldnot have been dissolved after the deposition ofthe overlying anhydrites (Holt and Powers,1988). The depositional margin of haliterepresents the location where dissolutionwould be found if it were present. Although

-159-

.G I

I

I

;( \

~ \ \(" '. :/, l'

"('," ," j)

.. .. ....(

.,

, :

(lilt

I 3 MILES

SCALE

EXPLANATION

• DRILL HOLES CON TOUR INTERVAL ~ 100 FEET

Figure"' 22. Isopach of overburden for the Culebra Dolomite Member (from Holt and others, inprep.).

-160-

,~

Q,'y

<:,\

~~:'...\'

I

it',

--- -

-­....-- .. --:

ZONE OFPOTENTIAL

DISSOLUTION - --IN LOWER RUSTLER _:-,

)

(

)

'-

_\

,-'

ZONE bFSALADO

DISSOLUTION

- - I

EXPLANATION

• DRILL HOLES

Figure 23. Distribution of Rustler and Salado dissolution near the WIPP site (from Holt andothers, in prep.).

-161-

limited, data from hydrologic testing near thiszone of potential dissolution do not show thehigh transmissivity values expected from anarea where dissolution has occurred.

Most of the deeper dissolution of Rustler andSalado evaporites in the vicinity of the WIPPsite occurred in response to the hydrologicregime of the ancestral Pecos River. Outcropsof ancestral Pecos River gravels indicate thatthe Cenozoic stream level was at least 650 fthigher than the present Pecos River and over200 ft higher than the equivalent freshwaterhead within the Culebra east of the WIPP site,where the lowest measured transmissivities arefound. The ancestral Pecos hydrologic systemwas capable of recharging the Rustler and theunderlying Salado. The dissolution patternwithin the Salado and Rustler near the WIPPsite originated in response to this hydrologicsystem. Dissolution progressed northwardfrom the ancestral Pecos system.

The modern Pecos River is not hydrologicallycapable of recharging the Rustler or SaladoFonnations, so this dissolution pattern is notdeveloping. Flow directions within NashDraw are to the south toward the modernPecos River (Mercer, 1983), and it issuspected that flow within the Rustler/Saladodissolution residuum is to the south orsoutheast. Salado halite continues to dissolvewithin Nash Draw and, possibly, south of theWIPP site in response to modern recharge inand around Nash Draw. Salado dissolutioncannot progress northward without rechargefrom the south.

The primary controls on Culebratransmissivity variations are unloading, Saladohalite dissolution, and Rustler halitedissolution (Figure 24). East of, and acrossmost of, the WIPP site area, unloading is theprimary cause of fracturing within theCulebra. As overburden decreases to thewest, fractures increase. Within zones whichpenetrate the northern and southern margins ofthe WIPP site, possible Rustler dissolutionmay increase the degree of fracturing withinthe Culebra. Minor collapse and stress reliefdue to Rustler dissolution locally wouldenhance the fracturing. West and south of theWIPP site, Salado dissolution caused moreextensive fracturing within the Culebra and isthe dominant cause of fracturing. Fracturing

generated by Salado dissolution stronglyoverprints fracturing of different origin.

Culebra ConceptualHydrogeological Model

Holt and others (in prep.) synthesizedhydrologic, geochemical, and geologic datainto a conceptual hydrogeological model of theCulebra summarized here.

The present hydrologic conditions within theCulebra are a relic of a previous hydrologicsystem. Modern flow directions are notconsistent with the distribution of soluteswithin Culebra groundwater (Ramey, 1985),and the Culebra is not actively being rechargedacross the WIPP site area (Lambert andHarvey, 1987). Isotope studies of the Culebragroundwater near the WIPP site indicate thatthe water has been isolated from theatmosphere for at least 12 ka (Lambert, 1987;Lambert and Carter, 1987). Culebratransmissivity varies over six orders ofmagnitude across the WIPP site area (Mercer,1983). This variation can be attributed to theinteraction of several processes includingunloading, Salado dissolution, and localizeddissolution within the Rustler. Theseprocesses, particularly dissolution of Saladoand possibly Rustler halite, were driven by thehydrologic regime of the Cenozoic ancestralPecos River. As dissolution and unloadingprogressed, the higher Rustler transmissivitiesshifted to the north and east.

In summary, groundwater from the ancestralPecos River hydrologic system accessed,dissolved, and recharged the Permianevaporites near the WIPP site. Flow throughthe evaporite section was dominantly northand east across the WIPP site area. Surfacerunoff increased, and westerly-drainingtributaries progressively removed overburdenfrom the evaporite section, including theCulebra. Strain associated with differentialunloading provided pathways for groundwaterinto existing gypsum-filled fracture networksand generated new fractures. Along themargins of the river, Salado halite dissolvedand the overlying rocks collapsed to generateadditional fractures and penneability in theCulebra. Where halite was dissolved from theRustler, fractures were more abundant andcontributed to a higher transmissivity. As theregional base level dropped, flow in the

-162-

•

UPPER' RUSTLERHALITE'DISSOLUTION

CO FIRMED

'I', '

UNLOADING

UPPER RUSTLERHALITE; DISSOLUTION

POSSIBLE

UPPER RUSTLERHALITE DISSOLUTION

POS IBLE

"---'--'1' ,

l: ""'(;.,j

'''I J

___~---1-LOWER RUSTLERHALITE DISSOLUTION

POSSIBLE

1 l'r

'l ','

\•

•

SCALE

3 MILES

" ~ '\ ,', fJf,I' ,1/

H 'll\~ 1;"1 1-

fl ~~ (~ ~ "- ;, ""

(,

'\.'-1\ )

l'1\

•-....' SALADO"-

)DISSOLUTION

//

(

)'\\

/ <-~()

;/V,,\ fc ' •

/-y.?~

/~/

.\II

; ~I

'\•~ -

EXPLANATION

• DRILL HOLES

Figure 24. Primary controls on Culebra transmissivity (from Holt and others, in prep.).

-163-

Culebra reversed, and the system began todrain back out toward its earlier rechargeareas. Even if permeability is continuing todevelop within the Culebra because Culebragroundwater may be undersaturated withrespect to gypsum, the hydrologic system isnot likely to change rapidly across the WIPParea without a significant change in regionalbase level.

Six hydrogeologic zones are identified withinthe Culebra in the vicinity of the WIPP site(Figures 25, 26). The zones were delineatedon the basis of transmissivity, amount ofoverburden, degree of hydraulic confinement,presence or absence of gypsum pore-fillings,and dissolution of underlying Salado halite(Figure 11). These zones progressed eastwardin response to progressive unloading anddissolution driven by the ancestral Pecoshydrologic system. The hydrologic conditionswhich allowed this process to take place in theCenozoic are no longer active.

Tamarisk Member

The Tamarisk Member of the RustlerFormation (Figure 12) consists of a mudstone(M-3/H-3), laterally equivalent to halite,sandwiched between two anhydrites (A-2 andA-3) (Holt and Powers, 1988). The anhydriteunits act as confining beds (Confining Beds 3and 4 of Holt and others, in prep.), while themudstone (Water-Producing Unit 3 [WPU-3])is the least productive of the Rustler water­producing units. Halite (H-3) is the dominantrock type in this stratigraphic position east ofthe WIPP site (Holt and Powers, 1988).

Less is known about the hydraulic propertiesof the Tamarisk than about those of the otherRustler members. No hydraulic tests of theTamarisk anhydrites (A-2 and A-3) have everbeen performed because of apparent lowtransmissivity. Hydraulic tests of theTamarisk claystone have been attempted atonly four locations: H-3b3 (1984 unpublishedfield notes), DOE-2 (Beauheim, 1986), H-14(Beauheim, 1987), and H-16 (Beauheim,1987). Drillstem and/or pulse tests failed at allfour locations, as the transmissivities wereapparently too low to measure over a period ofseveral days. Based on similar testsperformed sU,c,cessfully on the unnamed lowermember at H-16, Beauheim (1987) concluded

that the transmISSIVIty of the Tamariskclaystone is likely 10-5 ft2/day or less. Thetransmissivity of the Tamarisk claystone mightbe expected to be greater west of the WIPPsite in Nash Draw where dissolution of theupper Salado and Rustler has occurred, but nodirect measurements have been made in thisarea. The transmissivities of the Tamariskanhydrites in Nash Draw have probably beenincreased by dissolution and subsidence­induced fracturing.

No measurements of the undisturbed hydraulichead within the Tamarisk claystone have everbeen made, due to low transmissivities. Fluidpressures have been measured within theTamarisk interval of H-16 since August 1987(Stensrud and others, 1988a, 1988b), but thepressure did not stabilize before the pressurebegan responding to the construction of thenearby air-intake shaft.

The Tamarisk claystone consists of mostlymudstone and claystone with minor amountsof siltstone (Holt and Powers, 1988).Gypsum occurs as fibrous fracture fillings,nodules, and displacive crystals (Holt andPowers, 1988). The nodules and displacivecrystals of gypsum are primary and attest tothe lack of circulating groundwaters since thePermian (Holt and others, in prep.). Thefibrous fracture-filling gypsum may havedeveloped without an external source of fluidas the dewatering of the clays within the unitcould provide sufficient fluid for the relocationof sulfate. Holt and Powers (1988) reportcollapse of A-3 in core from boreholes H-11and H-3 and attribute this collapse todissolution of small amounts of halite fromthis interval. This suggests at least atemporary vertical hydraulic connectionbetween the M-3/H-3 interval and other water­bearing units in the Rustler. The timing andnature of this possible connection cannot bedetermined with the limited data available.

Magenta Dolomite Member

Holt and others (in prep.) include the Magentawithin their hydrostratigraphic subdivision ofthe Rustler as Water Producing Unit 4 (WPU­4). The Magenta is confined by A-3 and A-4of Holt and Powers (1988) which areequivalent to Confining Beds 4 and 5 of Holtand others (in prep.). The Magenta is a fine-

-164-

i'I"__ I

. :,I

•

•

.~ /

•

! ., .~ . ,r; ~,

1 r i', ~ It r

3 MILES

SCALE

EXPLANATION

• DRILL HOLES

Figure 25. Hydrogeologic zones defined for the Culebra Dolomite Member (from Holt JnJothers, in prep.). Zone characteristics are listed in Table 1.

-165-

CHARACTERISTICS OF CULEBRA HYDROGEOLOGIC ZONES

'ZONE 1 • Confined ZONE 4 • Confined• Deeply Buried • Overburden Greatly Decreased• Few Fractures, Most Gysum • Abundant Open FracturesFilled • Vugs Open/Collapsed

• Vugs Filled with Anhydrite • Modern Transmissivity• Modern Transmissivity < 10-100 ft2 /day

< 0.01 ft 2/dayZONES • Confined, Leaky

ZONE 2 • Confined • Overburden Usually Small• Overburden Slightly Decreased • Dissolution of Underlying Halite

I

• More Fractures, Most Gypsum Increases Fracturing.......0\0\ Filled • Fractures Very Abundant, OpenI

• Anhydrite Vug Fillings Altering to • Vugs Open/CollapsedGysum • Modern Transmissivity

• Modern Transmissivity 10-1000 ft2 /day0.01-10 ft 2/day

ZONE 6 • UnconfinedZONE 3 • Confined • Overburden Minimal

• Overburden Significantly • Collapsed due to SaladoDecreased Dissolution

• Abundant Fractures, Some • No GysumGysum Filled • Fractures and Breccia Abundant

• Vugs Open or Filled with Gysum • Vugs Open/Collapsed• Modern Transmissivity • Modern Transmissivity

< 1->10 ft 2/day >1000 ft2 /day

Table 1. Characteristics of Culebra Hydrogeologic Zones.

WEST ZONES PROGRA DE EAST DUE TOPROGRESSIVE UNLOADING/ EROSION AND DISSOLUTION

EAST

ZONE 6 ZONE 5 ZONE 4 ZONE 3 ZONE 2 ZONE 1

FORMATION

. . . .. ..~:-: ~4 .

~~... ~.I.~~ ':::::::::z::::::::............ . . . . . . . ........... . . . . . .......

.h.--'4Il-

fL

IilJs"'l.1:­0;;11 f:0

ll~..q""Otv

---............

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .• • • • • • • • • • • t •••••••••••••••••••••••••••••••••••••••••

I~

0\.......,J

I

~ GYPSUMFILLEDFRACTURES

ji i OPEN1'1 FRACTURES

EXPLANATION

OOPENVUG e~I~~~M

VUGO

. ANHYDRITEFILLEDVUG

Figure 26. Conceptual cross-section of the Culebra hydrogeologic zones showing the distribution of features (from Holt and others, inprep.).

grained, gypsiferous, arenaceous dolomite(Holt and Powers, 1988).

Hydraulic tests of the Magenta dolomite at 16locations (Mercer, 1983; Beauheim, 1986,1987) yield transmissivity values generallyle~s than or equal to 0.1 ft2/day (Figure 27).HIgher values of transmissivity, 0.3 and 1.0ft 2/day, are found at H-6a and H-9arespectively. The two highest values ofMagenta tr~smissivity, 375 and 53 ft2/day,are found In Nash Draw at WIPP-25 andWIPP-27, respectively, where dissolution ofthe upper Salado has caused collapse andfracturing of the overlying Rustler.

Between 1979 and 1981, stabilized waterlevels were measured in 13 wells completed tothe Magenta (Richey, 1987b). The elevationsof these stabilized water levels can becompared to estimated freshwater-headelevations (Figure 28) calculated usingborehole-fluid-density (or specific-gravity)data (Mercer and others, 1981; Dennehy andMercer, 1982; Mercer, 1983; Lambert andRobinson, 1984; Richey, 1986, 1987a). Thecontours of the freshwater heads indicate asouthwesterly flow direction within theMagenta in the northern portion of NashDraw, and a generally westward flowdirection toward Nash Draw over the rest ofthe region. Holt and others (in prep.)concluded that the flow directions of the!'1agenta (Figure 28) were not significantlyInfluenced by fluid density variations.

The hydraulic head at H-6a (Figure 28) seemsanomalously low compared to that atsurrounding wells, and it causes a flexure inthe contour lines across the northern portion ofthe WIPP site. The low head at H-6a may berelated to drainage into dissolution cavities inthe upper Rustler, such as those encounteredat WIPP-33 (Sandia and USGS, 1981) abouta half n:tile southwest of H-6. If so, it may bea relatIvely localized perturbation in theMagenta potentiometric surface, and may notaffect heads as far to the east as is indicated onFigure 28.

Freshwater heads (Figure 28) at DOE-2, H­14, and H-16 were estimated from pressuremeasurements. The pressures measured atDOE-2 and --H-14 (Beauheim, 1986, 1987)represent upper bounds on the stabilized

formation pressures, while the pressuremeasured at H-16 reflects drainage from theMagenta into the nearby WIpp shafts since1981 (Beauheim, 1987). With these caveats,the estimated freshwater heads at DOE-2, H­14, and H-16 agree qualitatively with thepattern of freshwater heads derived from thewater-level data.

The Magenta was originally deposited asclastic gypsum and carbonate grains in welldeveloped strata (Holt and Powers, 1988).Upon shallow burial, the gypsum clasticcomponent was at least partly replaced byanhydrite and sulfate cements developedbetween clastic grains, probably asovergrowths on existing sulfate (Holt andPowers, 1988). The anhydrite has since beenreplaced by gypsum, and most of the clastictexture within the sulfate component of theMagenta has been erased (Holt and Powers,1988). Porosity within the Magenta isprimarily intergranular and formed by thedissolution of gypsum (Holt and others, inprep.). In those areas which originallycontained abundant clastic gypsum, small?pen vugs have developed. Fracture porosityIS less common, and the fractures are oftenfilled with gypsum. Disruption and fracturingfrom unloading and dissolution haveapparently been less effective in the Magentathan in the Culebra.

Forty-niner Member

The Forty-niner Member of the RustlerFormation (Figure 12) is composed of twoanhydrite and/or gypsum units (A-4 and A-5o.f Holt and Powers, 1988) separated by aSIltstone, mudstone, and claystone interbed(M-4). The anhydrite/gypsum units act asconfining beds (Confining Beds 5 and 6 of~olt and others, in prep.), while the claystoneIS ~ water-producing unit (Water-ProducingUnlt 5). M-4 is stratigraphically equivalent tohalite (H-4) east of the WIPP site.

M-4 has been hydraulically tested at only fourlocations: DOE-2 (Beauheim, 1986), H-3d(Beauheim and others, in prep.), H-14(Beauheim, 1987), and H-16 (Beauheim,1987). Transmissivities reported for theclaystone are 2.5 x 10-3 to 1.1 x 10-2 ft2/day atDOE-2, 3.9 x 10-3 ft2/day at H-3d, 3.0 x 10-2

to 7.1 x 10-2 ft2/day at H-14, and 5.0 x 10-3 to

-168-

f+--WIPP-SITE80UNDARY

~H-5s.01

eDOE-2 0001

eWIPP-30. 0004

.H-4a.0.06

~"

.WIPP-27.53..~~/~

.,~,

"1?."~

'-f

~~,~

..~"~

\~~~

'fII,1'~

.\~,~

~" H-6a·03WIPP-25, 375.~,.

~,>...A 0 '*"~'" A'... H-16.003.t-- ~----""':":7'";:""'-------+:-:-H-~2a~.';:-OO~j"'=-:·H'e_,,-.-00-5---~-------~1

~ "11" t/" eH-3,01

" '~rr'~<. .....H-14,O.OO6..~'"~

N

J

H-9s. 10.

H-l0a.001.

eH-es.OOO6

LEGEND~WELLTESTED

• H-3. 0.1L MAGENTA TRANSMISSIVITY, ft 2/day

a 2 3

SCALE

4 mi

Figure 27. Transmissivity of the Magenta Dolomite Member (from Holt and others, in prep.).

-169-

~.....,..

~

~,

~/',

N

\\

\ 3100H·1Da -., . 3216

\~<6\

\(»~

-l<J'

\\\\\\\\\\\~

\ <J'

\<-bIf;

\\

3028.H-Se. 3029,

" "W~IPP28 ~::~5;

• WIPP-27 307~2 ~ ~308 ~ •

~ I.U~ ~. ~\\.- 0\

~\\. ,11\11 ' <)

~ ~

~ 0f f- \~ ,~ C/)

~\ 3056 I~\, H·6a. 3ci58_--t--+-.z.-_~-9-", H-5a.~

WIPP.25~. ~ 0 I 31~. 3056 ~

~\' ~~\\. ~ "

.~\h,;\ ~

~t \

~/t. \....~/' "

"lfl" •~

.,,~.

"/-i:.~~

.//~

,~~,,\~

~~ '\\,\,

.\~

"11~

J:...*'1'1 '1 ' "~\"

LEGEND

fJ:=WELL TESTED

.r--MEASURED WATER-LEVELT ELEVATION, ft amst

.H-2a 3149, 3152

! ESTIMATED FRESHWATER-HEADl--ELEVATION, ft amsl

2 3; !

SCALE

4 mi

I

-3125- FRESHWATER-HEAD ELEVATION CONTOUR

CONTOUR INTERVAL= 25 It

Figure 28. Measured water level and estimated freshwater head elevation, Magenta DolomiteMember (from Holt and others, in prep.).

-170-

5.6 x 10-3 ft2/day at H-16. The Forty-ninerclaystone might be assumed to have highertransmissivities in Nash Draw, but no directmeasurements have been made in that area.Transmissivity of the unit is probably less eastof the WIPP site where halite (H-4) is present.

The transmissivities of the anhydrites are verylow and not measurable (Beauheim, 1987).Where present in Nash Draw, the Forty-nineranhydrites may have very high transmissivitiesas a result of dissolution and subsidence­induced fracturing.

Hydraulic heads of the Forty-niner claystonehave been measured at wells H-3d, H-14, H­16, and DOE-2. Holt and others (in prep.)summarize these observations and estimatestatic fluid pressures at the midpoint of theclaystone. The measurement from DOE-2 hadthe greatest uncertainty. The original datafrom DOE-2 (Beauheim, 1986) indicate that a"static" pressure 12 psi lower than thatreported by Beauheim (1987) is a moreaccurate upper bound on the hydraulic head.Static fluid-pressure estimates for the Forty­niner claystone at H-3d, H-14, H-16, andDOE-2 have been converted to freshwaterheads (Figure 29). Firm conclusions cannotbe drawn from so few data, but the data areconsistent with southwesterly to westerly flowtowards Nash Draw. The relatively low headat H-16 may be related more to localizeddrainage into the WIPP shafts than to theundisturbed regional head distribution(Beauheim, 1987).

The M-4 unit (WPU-5) consists of siltstone,claystone, and locally some sandstone, all ofwhich are cemented with halite (Holt andPowers, 1988). Porosity within M-4 isprincipally intergranular, resulting from thedissolution of halite cements.

Summary of Rustler Hydraulic­Head Relationships

The hydraulic-head distributions for theunnamed lower member/upper Salado,Magenta, and Forty-niner claystone (Figures14, 28, and 29) all indicate westerly tosouthwesterly groundwater flow. In contrast,flow within the more transmissive Culebraappears to be largely southerly (Figure 16).These data are consistent with the

paleorecharge/modern discharge hypothesis ofLambert and Harvey (1987).

Vertical gradients within the Rustler reflect arecharge-discharge cycle and are consistentwith Lambert and Harvey's (1987) hypothesisof late-Pleistocene recharge and currentdischarge (Holt and others, in prep.). Avertical east-west cross section through thecenter of the WIPP site shows the directionsof the vertical hydraulic gradients betweenmembers of the Rustler (Figure 30). Whereno dissolution of either upper Salado orunnamed lower member halite has occurred,from about the west-central portion of theWIPP site to the east, hydraulic gradients fromthe unnamed member and Magenta indicatepotential fluid flow from both above andbelow into the Culebra. The Culebra, themost transmissive part of the Rustler in thisvicinity, drains towards Nash Draw faster thanthe other water-producing units. Thegradients toward the Culebra do not imply,however, that vertical leakage between units issignificant (Holt and others, in prep.).Rather, the high hydraulic heads in the low­permeability Magenta and unnamed memberdemonstrate the effectiveness of the confiningbeds separating them from the Culebra.

The western portion of the WIPP site appearsto be a transition region in which verticalhydraulic gradients change. Where the upperSalado has been dissolved, the Magenta headdecreases until the unit dewaters, andgradients are downward from the Culebra tothe unnamed member/upper Salado. Thechanges in transmissivities caused by thedissolution increased the internal drainagefrom each of the members to the west andsouthwest. Dissolution of the upper Saladocombined with Rustler subsidence to fracturethe Rustler confining beds, allowing increasedflow between Rustler members.

Vertical hydraulic gradients between theMagenta and the Forty-niner claystone areupward in the central portion of the WIPP site,based on measurements made in holes DOE-2,H-3d, H-14, and H-16 (Beauheim, 1987).The two water-producing units are separatedby an effective confining bed (A-4) in thisregion, which likely prevents significantleakage. The origin of the head differencebetween the Magenta and Forty-ninerclaystone is unclear, but the relative heads

-171-

WIPP·280

~

o

o WIPP-'~oH-'8 oWIPP-'§

8 ~:~~:;2H-16. 3107. gWIPp·2'

H-20 0" OH-'SH-' ERDA-9eH-3d,3124

H-14~<3111 ODOE-'

OBELCO

....EC-70

o P,'8

--WIPP-SITEBOUNDARY

OH·"

.DOE-2,<3163H·SO

o WIPP-'3

OH-4coCABIN BABY

oP-'7 OH-17

p·,S

°H-6

o WIPP-27

~ 0,~\' r,..

~\.' v~ ~

WIPP-2So 'f "~" ~

~\\, ~,j\k.~ '\

~t V op.\4

~WIPP·26

~'"'!fl"~

~..... 0-268~,..,. 0

"

~

';~,

';~

.,~

"/~

OW1PP·29

.,~

,~~

"'~.,\\1.\~

...~~

I\"~ -<-,,,,.."".

N

~/,~7/;,

..;;~

OH-7 ~~\",:::

~\,..~.~ '!lIk, .~\' ~

~\\\'" ,t;;,

o H·'2

H-'Do

J~-

~...........

~

~,

~/',J:~

o USGS-1

OH-9

OENGLE

OH-8b

LEGEND

.----WELL TESTED

• H-16. 3107... ESTIMATED FRESHWATERL-HEAD. ft amsl

2 3, ;

SCALE

4 miI

o OTHER OBSERVATION WELLS

Figure 29. Estimated freshwater heads for the claystone of the Forty-niner Member (from Holtand others, in prep.).

-172-

A4500 ~

Westt+?I

4000 II?

A'

EastDirectIOn of Hydraulic Gradient from Well Data

DirectIOn of Hydraulic Gradient Inferred from Regional Geology

Direction of Hydraulic GradIent Unknown

Q)

>Q)

roQ)enQ)

>0..cro......Q)Q)-I "-'

~

W-.......lVJ

0I

::J~

~-.J«

3500

3000

2500

I?I'

Dissolution of Upper Salado

Nash Draw

I WIPP SITE .1I~ II II III

?Halite inUnnamed Lower Member (M-1)

Halite in Tamarisk

Top ofRustlerMagenta

Culebra

Top ofSalado salt

2000o 5 10

Halite in Forty-Niner

15

APPROXIMATE DISTANCE FROM CENTER OF NASH DRAW (miles)

h~lIre '0 Vertlcal flow potential in the Rustler Fonnation (from Holt and others, in prep.).

currently prevent overlying units fromrecharging the Magenta. No data are availableto evaluate hydraulic gradients between theMagenta and Forty-niner claystone in theregion surrounding the WIPP site, such as atthe J.C. Mills Ranch to the south, whereportions of the Dewey Lake Formation aresaturated (Mercer, 1983).

The hydraulic gradients within and betweenthe individual water-producing units of theRustler are consistent with Lambert andHarvey's (1987) hypothesis that the Rustlerflow systems are not at steady-state, but areinstead draining, following a late-Pleistocenerecharge event (Holt and others, in prep.).The most transmissive unit, the Culebra, hasdrained faster than the other units, leaving itrelatively underpressurized at the WIPP siteand east of it. The other Rustler units exhibithydraulic gradients towards the drainageregion represented by Nash Draw. West ofthe WIPP site, where the upper Saladodissolved and caused Rustler subsidence andfracturing, it is likely that all of the Rustlermembers hydraulically communicate.

Conceptual HydrogeologicalModel of the Rustler

Holt and others (in prep.) proposed thefollowing conceptual model for thedevelopment of the hydrologic zones withinthe Rustler. In the WIPP region, thehydrogeologic conditions in the Rustlerdeveloped concurrently with the ancestralPecos hydrologic regime. The erosionalpattern generated by the ancestral Pecos Riverand its tributaries was superimposed over theexisting erosional surface. This erosionbrought the evaporites of the Rustler andSalado close to the surface, allowing rechargefrom the ancestral Pecos and meteoric watersto dissolve the evaporites. Groundwatercirculated and evaporites dissolved in areasadjacent to the ancestral Pecos River and itstributaries. Collapse enhanced thetransmissivity within the Rustler.

Major tributaries to the ancestral Pecos Riveralso dissolved evaporites, as erosion and laterdissolution preferentially exposed the Rustlerand Salado to circulating groundwater. Localsolution features collapsed along thesedrainages and later coalesced into karst

valleys. The hydrologic regime within theevaporite section around each of thesetributaries was dominated by these karstprocesses. Nash Draw was the closest suchtributary. During the Cenozoic, it acted as arecharge point and, as climatic conditionsvaried, a possible discharge point for fluidsmoving within the Rustler.

Ancestral Pecos River deposits can beconsidered a record of paleohydraulic headlevels. West and southwest of the WIPP(Figure 23), these deposits are higher than themodem head levels within any of the Rustlerunits when the effect of subsidence due todissolution is removed. This relationshipsuggests that the ancestral Pecos River and itstributaries within karst valleys were capable ofrecharging Rustler hydrologic units during theCenozoic.

The hydrologic systems within the Rustleracross the site area were mostly passive, as noactive discharge point existed. Additionalrecharge to the low-permeability parts of theRustler only occurred when new porositydeveloped in response to fracturing caused byunloading and dissolution of evaporitecements, fracture fillings, and sedimentaryfeatures. The Rustler hydrologic units weremost active near the ancestral Pecos River andits tributary hydrologic system, wheredissolution and collapse enhanced the porosityand permeability within the Rustler. Passiveconditions existed further away from therecharge areas, where conditions within theRustler remained at or near steady state, andwhere discharge was not possible or veryslow. The hydrologic conditions in the areabetween those clearly active and passivehydrologic zones varied with permeability,porosity, recharge rates, and discharge ratesdecreasing away from the active zones. Theactive hydrologic zones encroached upon thepassive hydrologic zones as unloading,erosion, and dissolution migrated laterallyaway from recharge areas.

Variations in climate, the location of theancestral Pecos River and tributary channels,and local geologic conditions controlled therate of advance and the locations of the activezones. As a consequence, the hydraulic headpotentials within the Rustler hydrologic unitsvaried such that flow direction reversed.Paleoflow reversals dissolved and removed

-174-

evaporite minerals from those passive areas ofthe Rustler and increased transmissivity andpotential for further discharge. As moderntransmissivity variations within the hydrologicunits reflect the extent of active Cenozoiczones, paleoflow directions within activeCenozoic hydrologic zones must have beenparallel to the modern transmissivitydistribution. The principal direction of flowhas not been precisely identified.

The hydrologic units within the Rustler acrossthe WIPP site are not being recharged today,and the flow directions indicate that Rustlerhydrologic units are now draining. Therefore,modern transmissivities of these units areinterpreted to reflect the maximumencroachment of Cenozoic active areas intopassive areas. The modern transmissivitieswithin Rustler hydrologic units in the vicinityof the WIPP site generally decrease away fromNash Draw, a probable Cenozoic rechargearea. Modern transmissivities appear toincrease where Cenozoic dissolution andcollapse occurred within the Rustler andSalado. In those hydrologic units subject tofracturing during unloading (principally theCulebra, WPU-2), modern transmissivitygenerally is related to the amount ofoverburden removed by erosion. This is truein the highly transmissive zone south of theWIPP site and east of Nash Draw. Moderntransmissivities may be lower than predicted inthose areas where late-stage secondary porefillings are present. At the present time, onlyone such location has been tentativelyidentified in the Rustler. It occurs within theCulebra Dolomite Member (WPU-2) west ofthe WIPP site.

Dewey LakeFormationThe Dewey Lake consists almost entirely ofmudstone, claystone, siltstone, andinterbedded sandstone. The lower DeweyLake is characterized by locally abundantgypsum-filled fractures. The upper DeweyLake is cemented with carbonate, and it ismoist within the air-intake shaft at the WIPP(Holt and Powers, 1990). Some fractures inthis zone are open, and carbonate-filledfractures become more abundant downward.

Lower in the Dewey Lake, an abrupt changeof cement type was noted by Holt and Powers(1990) from carbonate to a much harder,unidentified material. Coincident with thechange in cement, fractures became filled withfibrous gypsum, and no moisture was evident.

In the northern and central portions of theWIPP site, no water has been reported in theDewey Lake in WIPP drillholes, although azone of lost circulation associated withfractures is commonly encounteredapproximately 200 to 250 ft below groundsurface (e.g., Mercer and others, 1987).Water has been encountered within the DeweyLake at several drillholes near the southernWIPP site boundary, and several stock wellssouth of the WIPP site are possibly completedin a perched aquifer in the upper Dewey Lake(Mercer, 1983).

Dockum GroupThe Dockum Group is of Late Triassic age(Figure 1). The Dockum Group occurs as anerosional wedge that pinches out westwardalong a north-south line through the middle ofthe WIPP site (Powers and others, 1978).The Dockum Group consists of calcareousinterbedded sandstone, siltstone, andclaystone. It is moderately well indurated tosoft. It is not a hydrologically significant unitin the vicinity of the WIPP site, although it is asignificant aquifer east of the WIPP site(Mercer, 1983). The Dockum Group may berecharged by precipitation in the easternportion of the WIPP site (Mercer, 1983).

References CitedAnderson, R Y., and Powers, D W., 1978,

Salt anticlines in Castile-Saladoevaporite sequence, northern DelawareBasin, New Mexico, in Austin, G.S.,compiler, Geology and mineraldeposits of Ochoan rocks in DelawareBasin and adjacent areas: New MexicoBureau of Mines and MineralResources, Circular 159, p. 79-83.

-175-

Bachman, G.O., 1987, Karst in evaporites insoutheastern New Mexico: SAND86­7078, Sandia National Laboratories,Albuquerque, NM.

Barr, G.E., Lambert, S.J., and Carter, J.A.,1979, Uranium isotope disequilibriumin groundwaters of southeastern NewMexico and implications regardingage-dating of waters: IsotopeHydrology 1978, v. 2, InternationalAtomic Energy Agency, p. 645-660.

Beauheim, R.L., 1986, Hydraulic-testinterpretations for well DOE-2 at theWaste Isolation Pilot Plant (WIPP)site: SAND86-1364, Sandia NationalLaboratories, Albuquerque, NM.

Beauheim, R.L., 1987, Interpretations ofsingle-well hydraulic tests conducted atand near the Waste Isolation Pilot Plant(WIPP) site, 1983-1987: SAND87­0039, Sandia National Laboratories,Albuquerque, NM.

Beauheim, R.L., 1988, Scale effects in welltesting in fractured media, inProceedings, Fourth Canadian/American Conference onHydrogeology, Fluid Flow, HeatTransfer, and Mass Transport inFractured Rocks: National Water WellAssociation, p. 152-159.

Beauheim, R.L., Dale, T.F., and Pickens,J.F., in prep., Interpretations ofsingle-well hydraulic tests of theRustler Fonnation conducted in thevicinity of the Waste Isolation PilotPlant (WIPP) site, 1988-1989:SAND89-0869, Sandia NationalLaboratories, Albuquerque, NM.

Beauheim, R.L., Hassinger, B.W., andKlaiber, J.A., 1983, Basic data reportfor borehole Cabin Baby-1 deepeningand hydrologic testing: WTDS-TME­020, U.S. Department of Energy,Albuquerque, NM.

Beauheim, R.L., Saulnier, G.J., Jr., andAvis, J.D., in prep., Interpretation ofbrine-penneability tests of the SaladoFormation at the Waste Isolation Pilot

Plant site: first interim report:SAND90-0083, Sandia NationalLaboratories, Albuquerque, NM.

Berg, R.R., 1979, Reservoir sandstones ofthe Delaware Mountain Group,southeast New Mexico, inGuadalupian Delaware MountainGroup of West Texas and SoutheastNew Mexico: Pennian Basin SectionPub. 79-18, Society EconomicPaleontologists and Mineralogists, p.75-95.

Borns, D.J., and Shaffer, S-E., 1985,Regional well-log correlation in theNew Mexico portion of the DelawareBasin: SAND83-1798, SandiaNational Laboratories, Albuquerque,NM.

Borns, D.J., and Stonnont, J.C., 1988, Aninterim report on excavation effectstudies at the Waste Isolation PilotPlant: the delineation of the disturbedrock zone: SAND87-1375, SandiaNational Laboratories, Albuquerque,NM.

Borns, D.J., Barrows, L J., Powers, D.W.,and Snyder, R P., 1983, Defonnationof evaporites near the Waste IsolationPilot Plant (WIPP) site: SAND82­1069, Sandia National Laboratories,Albuquerque, NM.

Bredehoeft, J.D., 1988, Will salt repositoriesbe dry?: EOS, Transactions AmericanGeophysical Union, v. 69, no. 9.

Cauffman, T.L., LaVenue, A.M., andMcCord, J.P., 1990, Ground-waterflow modeling of the CulebraDolomite: Volume II - Data Base:SAND89-7068/2, Sandia NationalLaboratories, Albuquerque, NM.

Crawley, M.E., 1988, Hydrostatic pressureand fluid density distribution of theCulebra Dolomite Member of theRustler Fonnation near the WasteIsolation Pilot Plant, southeastern NM:DOE/WIPP 88-030, U.S. Departmentof Energy, Carlsbad, NM.

-176-

Davies, P.B., 1989, Variable-density groundwater flow and paleohydrology in theWaste Isolation Pilot Plant (WIPP)region, southeastern New Mexico:Open-File Report 88-490, U.S.Geological Survey.

Deal, D.E., and Case, J.B., 1987, Brinesampling and evaluation program,phase I report: DOE-WIPP-88-008,U.S. Department of Energy, Carlsbad,NM.

Dennehy, K.F., and Mercer, J.W., 1982,Results of hydrologic tests and water­chemistry analyses, wells H-SA, H­SB, and H-SC, at the proposed WasteIsolation Pilot Plant site, southeasternNew Mexico: Water ResourcesInvestigation 82-19, U.S. GeologicalSurvey, Albuquerque, NM.

Eanh Technology Corporation, 1988, Finalreport for time domain electromagnetic(TDEM) surveys at the WIPP site:SAND87-7144, Sandia NationalLaboratories, Albuquerque, NM.

Grauten, W.F., 1965, Fluid relationships inDelaware Mountain Sandstone, inFluids in Subsurface Environments:Memoir 4, American Association ofPetroleum Geologists, p. 294-308.

Haug, A., Kelley, V.A., LaVenue, A.M., andPickens, J.F., 1987, Modeling ofground-water flow in the CulebraDolomite at the Waste Isolation PilotPlant (WIPP) site: interim report:SAND86-7167, Sandia NationalLaboratories, Albuquerque, NM.

Hills, J.M., 1984, Sedimentation, tectonism,and hydrocarbon generation inDelaware Basin, west Texas andsoutheastern New Mexico: AmericanAssociation Petroleum GeologistsBulletin, v. 68, p. 250-267.

Hiss, W.L., 1976, Stratigraphy and ground­water hydrology of the Capitanaquifer, southeastern New Mexico andwest Texas [Ph.D. Thesis]: Boulder,University of Colorado.

Holt, R.M., and Powers, D.W., 1984,Geotechnical activities in the wastehandling shaft: WTSD-TME-038,U.S. Dept. of Energy, Carlsbad, NM.

Holt, R.M., and Powers, D.W., 1986,Geotechnical activities in the exhaustshaft: DOE/WIPP 86-008, U.S. Dept.of Energy, Carlsbad, NM.

Holt, R.M., and Powers, D.W., 1988, Faciesvariability and post-depositionalalteration within the Rustler Formationof the Waste Isolation Pilot Plant,southeastern New Mexico: DOE­WIPP-88-004, U.S. Department ofEnergy, Carlsbad, NM.

Holt, R.M., and Powers, D.W., 1990,Geologic mapping of the air intakeshaft at the Waste Isolation Pilot Plant:U.S. Department of Energy, Carlsbad,NM.

Holt, R.M., Powers, D.W., Beauheim, R.L.,and Crawley, M.E., 1988, Origin oftransmissivity variations in the CulebraDolomite near the Waste Isolation PilotPlant WIPP) site, southeastern NewMexico: Ground Water, v. 26, p.790.

Holt, R.M., Powers, D.W., Beauheim, R.L.,Crawley, M.E., and Abitz, R., inprep., Conceptual hydrogeologicalmodel of the Rustler Formation in thevicinity of the Waste Isolation PilotPlant site, southeastern New Mexico:SAND89-0862, Sandia NationalLaboratories, Albuquerque, NM.

Hudson, J., 1987, Preliminary pers. comm.to J.S. Stormont (Sandia NationalLaboratories) on USGS moisture anddensity measurements in bedded rocksalt at the WIPP site.

Jones, C.L., Cooley, M.E., and Bachman,G.O., 1973, Salt deposits of LosMedafios area, Eddy and LeaCounties, New Mexico: Open-FileReport 4339-7, U.S. GeologicalSurvey, 67 p.

-177-

Lambert, S.l., 1987, Stable-isotopic studiesof groundwaters in southeastern NewMexico: SAND85-1978C, SandiaNational Laboratories, Albuquerque,NM, in Chaturvedi, L., ed., TheRustler Formation at the WIPP site:EEG-34, Environmental EvaluationGroup, Santa Fe, NM.

Lambert, S.l., and Carter, J.A., 1987,Uranium-isotope systematics ingroundwaters of the RustlerFormation, northern Delaware Basin,southeastern New Mexico: 1.Principles and Preliminary Results:SAND87-0388, Sandia NationalLaboratories, Albuquerque, NM

Lambert, S.J., and Harvey, D.M., 1987,Stable-isotope geochemistry ofgroundwaters in the Delaware Basin ofsoutheastern New Mexico: 1:Principles and Preliminary Results:SAND87-0138, Sandia NationalLaboratories, Albuquerque, NM.

Lambert, S.J., and Robinson, K.L., 1984,Field geochemical studies ofgroundwaters in Nash Draw,southeastern New Mexico: SAND83­1122, Sandia National Laboratories,Albuquerque, NM.

Lang, W.B., 1938, Geology of the PecosRiver between Laguna Grande de laSal and Pierce Canyon: 12th and 13thBiennial Reports, New Mexico StateEngineer, p. 80-86.

Lappin, A.R., 1988, Summary of site­characterization studies conductedfrom 1983 through 1987 at the WasteIsolation Pilot Plant (WIPP) site,southeastern New Mexico: SAND88­0157, Sandia National Laboratories,Albuquerque, NM.

LaVenue, A.M., Haug, A., and Kelley, V.A.,1988, Numerical simulation ofground-water flow in the CulebraDolomite at the Waste Isolation PilotPlant (WIPP) site: second interimreport: SAND88-7002, SandiaNational Laboratories, Albuquerque,NM.-

LaVenue, A.M., Cauffman, T.L., andPickens, J .F. 1990, Ground-waterflow modeling of the CulebraDolomite: Volume I: modelcalibration: SAND89- 7068/1, SandiaNational Laboratories, Albuquerque,NM.

McNeal, R.P., 1965, Hydrodynamics of thePermian Basin, in Flu ids i nSubsurface Environments: Memoir 4,American Association PetroleumGeologists, p. 308-326.

Mercer, J.W., 1983, Geohydrology of theproposed Waste Isolation Pilot Plantsite, Los Medaiios area, southeasternNew Mexico: Water ResourcesInvestigation 83-4016, U.S.Geological Survey, Albuquerque,NM.

Mercer, J.W., 1987, Compilation ofhydrologic data from drilling theSalado and Castile Formations near theWaste Isolation Pilot Plant (WIPP) sitein southeastern new Mexico:SAND86-0954, Sandia NationalLaboratories, Albuquerque, NM.

Mercer, J.W., Davis, P., Dennehy, K.F., andGoetz, C.L., 1981, Results ofhydrologic tests and water-chemistryanalyses, wells H-4A, H-4B, and H­4C at the proposed Waste IsolationPilot Plant (WIPP) site in southeasternNew Mexico: Water ResourcesInvestigations 81-36, U.S. Geo­logical Survey, Albuquerque, NM.

Mercer, J.W., Beauheim, R.L., Snyder,R.P., and Fairer, G.M., 1987, Basicdata report for drilling and hydrologictesting of drillhole ooE-2 at the WasteIsolation Pilot Plant (WIPP) site:SAND86-0611, Sandia NationalLaboratories, Albuquerque, NM.

Nowak, E.J., McTigue, D.F., and Beraun,R., 1988. Brine inflow to WIPPdisposal rooms: data, modeling, andassessment: SAND88-0112, SandiaNational Laboratories, Albuquerque,NM.

-178-

Stein, C.L., and Krumhansl, J.L., 1986,Chemistry of brines in salt from theWaste Isolation Pilot Plant (WIPP) ,southeastern New Mexico: apreliminary investigation: SAND85­0897, Sandia National Laboratories,Albuquerque, NM.

Sandia National Laboratories and U.S.Geological Survey, 1981, Basic datareport for drillhole WIPP 33 (WasteIsolation Pilot Plant - WIPP):SAND80-2011, Sandia NationalLaboratories, Albuquerque, NM.

Siegel, M.D., Lambert, S.J., and Robinson,K.L., eds., 1990, Hydrogeochemicalstudies of the Rustler Formation andrelated rocks in the WIPP area,southeastern New Mexico: SAND88­0196, Sandia National Laboratories,Albuquerque, NM.

Stensrud, W.A., Bame, M.A., Lantz, K.D.,Cauffman, T.L., Palmer, J.B., andSaulnier, G.J., Jr., 1988a, WIPPhydrology program, Waste IsolationPilot Plant, southeastern New Mexico,hydrologic data report #6: SAND87­7166, Sandia National Laboratories,Albuquerque, NM.

Stensrud, W.A., Bame, M.A., Lantz, K.D.,Palmer, J.B., and Saulnier, G.J., JT.,1988b, WIPP hydrology program,Waste Isolation Pilot Plant,southeastern New Mexico, hydrologicdata report #7: SAND88-7014, SandiaNational Laboratories, Albuquerque,NM.

Pfeifer, M.C., 1987, Multicomponentunderground DC resistivity study atthe Waste Isolation Pilot Plant,southeastern New Mexico [M.S.Thesis T-3372]: Golden, ColoradoSchool of Mines.

Popielak, R.S., Beauheim, R.L., Black,S.R., Coons, W.E., Ellingson, C.T.,and Olsen, R.L., 1983, Brinereservoirs in the Castile Formation,southeastern New Mexico: TME-3153,U.S. Dept. of Energy, Albuquerque,NM.

Powers, D.W., Lambert, S.J., Shaffer, S-E.,Hill, L.R., and Weart, W.D., eds.,1978, Geological characterizationreport, Waste Isolation Pilot Plant(WIPP) site, southeastern NewMexico: SAND78-1596, v. I & II,Sandia Laboratories, Albuquerque,NM.

Ramey, D.S., 1985, Chemistry of RustlerFluids: EEG-31, EnvironmentalEvaluation Group, Santa Fe, NM.

Richey, S.F., 1986, Hydrologic-test datafrom wells at hydrologic-test pads H­7, H-8, H-9, and H-I0 near theproposed Waste Isolation Pilot Plantsite, southeastern New Mexico: Open­File Report 86-413, U.S. GeologicalSurvey, Albuquerque, NM.

Richey, S.F., 1987a, Preliminary hydrologicdata for wells tested in Nash Draw,near the proposed Waste Isolation PilotPlant site, southeastern New Mexico:Open-File Report 87-37, U.S.Geological Survey, Albuquerque,NM.

U.S. GeologicalAlbuquerque, N.M.

Survey,

Richey, S.F., 1987b, Water-level data fromwells in the vicinity of the WasteIsolation Pilot Plant, southeastern NewMexico: Open-File Report 87-120,

Westinghouse Electric Corporation, 1988,Geotechnical field data and analysisreport: DOE-WIPP-87-0l7, U.S.Dept. of Energy, Carlsbad, NM.

-179-