SAND82 -1069Unlimited Release

Printed March 1983

DistributionCategory UC-70

David J. BornsEarth Sciences Division 9731

Dennis W. PowersDepartment of Geological Science

University of Texas at EI PasoEI Paso, TX 79968

Lawrence J. BarrowsGround Motion and Seismic Division 7111

Sandia National LaboratoriesAlbuquerque, NM 87185

PO"TlOv.~ NOTICE ...it~~~. '--' Of T!!,S REPORT ARE 'UEGlBLE'"

,,8 berm reprQd I ' ....:

availaIJ5e copy to uceo. from the besfPossible avafiabil.p penmi the. brOades(

i~ j

Richard P. SnyderUS Geological Survey

Denver Federal CenterDenver, CO 80225

3

DISm!IlUnOU Of THIS DOCUMENT IS UNlIMiT/Il ~Vi

AbstractThe Delaware Basin is a broad asymmetric sedimentary trough in southeastern New Mexico and westTexas. Basin subsidence occurred from the Pennsylvanian into the Triassic. The basin also underwenttilting since the early Cenozoic.

Layered evaporite units of Ochoan age in the basin are 1000 m thick. These evaporites are divided intothree stratigraphic units (listed in order of increasing age): the Rustler Formation, the SaladoFormation. the Castile Formation. These units, especially the Castile, are deformed along portions ofthe margin of the Delaware Basin and in some areas internal to the basin. In the northern Delaware Ba­sin adjacent to the WIPP site, the term "Disturbed Zone" (DZ) has been applied to an area in which de­formation structures are found in boreholes and from which chaotic seismi~ reflection data wereobtained. The origin and timing of this deformation is considered important for the determination of

This document isPUBLICLY RELEASABLE

.tMyY ~~d4rr---Authorizing Officlal

Date: 1l/~;,' I 'U''' ,

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

DISCLAIMER Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

4

Abstract (cont)failure scenarios for the WIPP site. However, the defbrmation does not present a major hazard to theconstruction and operational stages of the facility.

Geophysical studies (borehole logging, seismic reflection, and detailed gravity) have covered the DZareas. Logs show vertical relief in the order of tens to hundreds of meters. Seismic profiles suggest ablocky structure with abrupt offsets and changes in dip between units. This chaotic structure occurswithin the Castile Formation, but with little or no deformation exhibited by the overlying andunderlying strata. Changes in seismic character, such as wiggle shapes, imply that unit thicknessesand/or acoustic properties vary. On the periphery of the DZ, seismic profiles indicate salt flowageflexures. Outside the DZ, the seismic time structures appear to be generated by lateral velocityvariations. The gravity field is dominated by anomalies originating in lateral density variations withinrelatively flat-lying strata. Low-amplitude, long-wavelength effects of DZ structures cannot be re­solved.

Through detailed core description, both meso- and microscopic, fold-styles in the DZ units areseparable into distinct stages and sequences of generation. An older sedimentary stage of folding isdistinguished from later tectonic folding by microfabric development and opposite senses of asymme­try. The tectonic folding and deformation displays a progressive development of fabrics ending in late­stage fracturing of competent anhydrite units. Petrofabrics reveal synkinematic growth of rotatedanhydrite porphyroblasts and stress shadow growth. Microboudinage is also common. Of the possibledeformation mechanism for anhydrite and halite, pressure solution appears the most applicable to theDZ. Therefore, an intergranular fluid plays an important role in facilitating deformation under thepressure, temperature, and stress histories of the region.

The following hypotheses of origin of deformation have been considered:

Gravity FounderingGravity SlidingGypsum DehydrationDissolutionDepositional Variations (e.g., thickness)

Of these, gravity foundering and sliding are considered the most probable causes of deformation.However, no hypothesis adequately answers why the deformation has a limited areal distribution. Apossible explanation would be areal variations in rock strength caused by variations of intergranularwater content. Age and timing of deformation are also crucial. Standard stratigraphic arguments basedon superposition may not apply to such a highly incompetent material as halite. Gravity founderingcould have happened at any time sincE;l deposition including the present; gravity sliding would probablyhave occurred since basin tilting began in the Cenozoic.

Deformation could be ongoing. However, the strain rates are such (~1O-16S-1) that deformation wouldprogress slowly relative to the facility's time frame of 2.5 x 105 yr. Deformation of Salado units would beminimal (< 10 m) or nonexistent, but within this time frame, upper anhydrite units of the Castile couldfracture and provide the volume for a brine reservoir. Such volumes would be small «1 %) and wouldrequire ~ 104 to 106 yr to develop. At these strain rates, fractures that connect the fractured anhydritesof the Castile with the middle Salado could not develop. Deformation should not directly jeopardize thefacility over th~ next 2.5 x 105 yr.

Contents1. Introduction 9

1.1 Organization of the Report 91.2 Purpose of the Report 91.3 Importance of the Deformation of Evaporites 91.4 Definition of the Disturbed Zone 10

2. General Site Geology 132.1 Introduction 132.2 Regional Geologic Setting 132.3 Stratigraphy and Lithology 14

2.3.1 Permian (pre-Ochoan) 142.3.2 Permian (Ochoan) 142.3.3 Triassic 142.3.4 Jurassic and Cretaceous 162.3.5 Tertiary 162.3.6 Quaternary 16

2.4 Structure and Thickness 162.4.1 Regional....................................................................................................................................... 162.4.2 Pre-Ochoan 162.4.3 Ochoan 162.4.4 Tertiary 442.4.5 Quaternary 44

2.5 Discussion and Conclusions 442,6 Selected References 51

3. Site Deformation 533.1 Borehole Data 533.2 Seismic Reflection 56

3.2.1 Data 563.2.2 Interpretation 653.2.3 Stratigraphic Correlation 703.2.4 Velocity Control......................................................................................................................... 71

3.3 Gravity 723.4 Detailed Core Description 82

3.4.1 Multiple Fold and Deformation Textures 823.4.2 Petrographic Description: Deformation Mechanism 83

4. Hypotheses of Origin 874.1 Gravity Foundering 874.2 Dissolution Mechanisms 884.3 Gravity Slides 894.4 Gypsum Dehydration 904.5 Depositional Processes 90

5. Discussion 935.1 Data Base 935.2 Geologic Structure 935.3 Age of Deformation 945.4 The Role of Pressure Solution 955.5 Rates of Structural Development 96

5

Contents (cont)6. Conclusions 103

6.1 Constraints on Age of Deformation 1036.2 Syndepositional to Closely Postdepositional Deformation 1036.3 Gravity Foundering 1046.4 Gravity Sliding 1046.5 Effects of Various Mechanisms on Site Suitability and Stability 104

References 105APPENDIX-Halokinetic Development of the Disturbed Zone

(by Larry Barrows, Sandia National Laboratories) 107

Figures1-11-22-12-22-3

2-4

2-52-62-72-82-92-102-112-122-132-142-152-162-172-182-192-202-212-222-232-242-252-262-272-282-292-30

6

Areal Extent of the Disturbed Zone, Northern Delaware Basin ..Present Configuration of Surface Control Zones at the WIPP Site ..Regional Setting, Northern Delaware Basin, Southeastern New Mexico .Generalized North-South Cross Section, Southeastern New Mexico ..Structure Contour Map, Top of Bell Canyon Formation, Northern Part of DelawareBasin .Index Map Showing Lines and Drill Holes Used in Cross Sections Shown onFigures 2-5 and 2-6 .North-South Cross Section, WIPP Site ..East-West Cross Section, WIPP Site ..Structure Contour Map, Top of Anhydrite Unit AI, Castile Formation ..Structure Contour Map, Top of Halite Unit HI, Castile Formation .Isopach Map, Halite Unit HI, Castile Formation .Structure Contour Map, Top of Anhydrite Unit All, Castile Formation .Structure Contour Map, Top of HaEte Unit HII, Castile Formation ..Isopach Map, Halite Unit HII, Castile Formation .Structure Contour Map, Top of Anhydrite Unit AlII, Castile Formation .Isopach Map, Anhydrite Unit AlII, Castile Formation ..Structure Contour Map, Base of Cowden Anhydrite Member, Salado Formation .Isopach Map, Base of Cowden Anhydrite Member to Top of Castile Formation ..Structure Contour Map, Top of Lower Unnamed Unit, Salado Formation ..Isopach Map, Lower Unit, Salado Formation ..Structure Contour Map, Top of McNutt Potash Zone, Salado Formation ..Isopach Map, McNutt Potash Zone, Salado Formation .Structure Contour Map, Top of Salado Formation .Isopach Map, Upper Unit, Salado Formation ..Location Map of 20 Drill Holes Used in Study of Salado Formation Marker Bed Intervals ..Structure Contour Map, Top of Rustler Formation .Isopach Map, Rustler Formation .Structure Contour Map, Top of Dewey Lake Red Beds ..Isopach Map, Dewey Lake Red Beds ..Structure Contour Map, Top of Santa Rosa Sandstone ..Isopach Map, Santa Rosa Sandstone ..Isopach Map, Gatufta Formation ..

10111315

17

202122232526272829303132333436373839404142434546474849

Figures (cont)3-1 Boreholes in the DZ That Penetrate the Castile Formation 533-2 Salado and Castile Formation Log Correlations in the:DZ 533-3 Salado and Castile Formation Log Correlations in the DZ 543-4 Acoustilog Correlation of the Salado Formation in AEC 8 and WIPP 11 553-5 Vibroseis Line Locations 573-6 Seismic Time Structure - Top of Castile Formation 583-7 Seismic Time Structure - Mid-Castile Formation 593-8 Seismic Time Structure - Top of Cherry Canyon Formation 603-9 Seismic Time Structure - Top of Bone Spring Formation 613-10 Seismic Isochron - Top Castile to Mid-Castile 623-11 Seismic Isochron - Mid-Castile to Cherry Canyon 633-12 Seismic Isochron - Cherry Canyon to Bone Spring 643-13 Seismic Line 77X5 663-14 Seismic Line 78GG3 673-15 Seismic Line 78GG19 683-16 Comparison of Interval Velocities Measured During Uphole Velocity Surveys -

ERDA 9, WIPP 12, WIPP 13 713-17 Compensated Densilog, Castile Formation, Borehole AEC 8 723-18 WIPP Gravity Survey, Detailed Areas, and Reconnaissance Profiles 743-19 WIPP Gravity Survey, Simple Bouguer Gravity Map 753-20 WIPP Gravity Survey, Simple Bouguer Gravity Map in Vicinity of Borehole

WIPP 14 763-21 Profile Across the Negative Gravity Anomaly at Borehole WIPP 14 773-22 That Portion of Seismic Line 77X2 That Intersects the Negative Gravity Anomaly......... 783-23 Gravity Survey Line G 803-24 Comparison of Intel"val Velocities Measured During Uphole Velocity Surveys -

WIPP 13, WIPP 34 803-25 Modeled Gravity Anomaly - Survey Line G............................................................................. 803-26 Deformation Styles in Laminated Carbonate Anhydrite Units of the Castile

(WIPP 13, 3727 ft) 823-27 Deformation Styles in Laminated Carbonate Anhydrite Units of the Castile

(WIPP 13, 3727 ft) 823-28 Deformation Styles in Laminated Carbonate Anhydrite Units of the Castile

(WIPP 13, 3729 ft) 833-29 Deformation Styles in Laminated Carbonate Anhydrite Units of the Castile

(WIPP 13, 3733 ft) 833-30 Synoptic Diagram of Log Stress vs Temperature Illustrates the Relative Strength at

10% Strain of Halite, Anhydrite, and Limestone Expected Under Geological Strain Rates.. 843-31 Rotated Anhydrite Neoblast; Synkinematic Growth Suggested by Helicitic

Calcite-Opaque Mineral Inclusion 853-32 Pressure Shadow Growth of Anhydrite Neoblast; Microboudinage of Calcite Laminae 865-1 Geometry of Boundary Development of a Structure Such as the WIPP 12 Anticline 101

7

18 W5670707997

98

99100101

Geologic Markers in Exploratory Test Holes .WIPP Site Seismic Surveys , .Stratigraphic Identification of Uppermost Seismic Horizon ..Tentative Stratigraphic Identification of Lower Three Seismic Horizons .Measured Velocities and Calculated WIPP 11 Depth-Time Tie .Strain Rates as Related to the Deformation Mechanism .The Effects on Comparative Strain Rates of Thermal and Stress DifferencesBetween WIPP Site and Gulf Coast Salt Domes .Comparative Strain Rate Calculations for Pressure Solution and Dislocation Climb @(J = 0.5 MPa and T = 308 K (35°C) ..Time Required for Developing WIPP 12-Type Anticline at 1% Volumetric Strain .Horizontal Rate Calculations for Growth of an Anticline Edge .

5-3

Tables2-13-13-23-33-45-15-2

5-45-5

8

Deformation of Evaporites Near theWaste Isolation Pilot Plant (WIPP) Site

1. Introduction

1. 1 Organization of theReport

This report is about a structural complex in theCastile and lowermost Salado Formations in thenorthern part of the WIPP site. The report is orga­nized to accomplish several purposes. The Introduc­tion provides a common ground for a definition of theDisturbed Zone (DZ) and points out some of its his­torical development. General Site Geology providesthe borehole data and the interpretation of the USGeological Survey (USGS). Site Deformation reviewsthe site structures that are based on borehole, geo­physical, and petrographic data. Hypotheses of Origindiscusses in detail the proposed mechanisms to ac­count for the DZ. (The gravity-foundering hypothesisis described in greater detail in the Appendix.) TheDiscussion probes some of the strengths and weak­nesses in data and hypotheses. In the Conclusions, wepresent the preferred hypotheses in the judgment ofthe Sandia authors and discuss these hypotheses interms of their relevance to site suitability.

1.2 Purpose of the ReportThis report summarizes work to date on the DZ

and related evaporite structures in the Delaware Ba­sin of southeast New Mexico. The deformation associ­ated with the DZ has been characterized by borehole,geophysical, and petrographic studies. The bulk ofthis work has been completed.

Various hypothetical mechanisms have been pro­posed to account for the observed deformation associ­ated with the DZ:

• Gravity foundering of denser anhydrite mem­bers of the Castile through the less dense halitemembers

• Dissolution• Gravity slide associated with basin tilting• Gypsum dehydration• Depositional variations in thickness

Investigations have included petrographic andgeochemical characterization of cores, interpretationof gravity antl level-line surveys, the relationship ofstructures to brine encounters in the Castile Forma­tion, and model studies of deformation mechanisms.Some of this material is contained here; all will becovered in reports when the investigations are com­plete.

1.3 Importance of theDeformation of Evaporites

Existing geologic structures at the WIPP site donot present an obstacle to facility development. How­ever, the development of reliable failure scenarios forthe next 250 000 yr requires information on the char­acteristics of the deformation, its age, mode of origin,and relation to fluids.

9

1.4 Definition of theDisturbed Zone

The evaporite beds are deformed around portionsof the margin of the Delaware Basin and in some areasin the middle of the Basin. Deformation features inthe northern part of the WIPP site have been lumpedtogether under the term "Disturbed Zone" (DZ). TheDZ is now delineated on the combined basis of struc­ture exhibited in boreholes and by the chaotic seismicreflection data in the northern part of the site (seeFigure 1-1).

• Structures on which the DZ is defined are thick­ened or thinned halite units of the Castile andanti- or synforms in anhydrite units of the Cas­tile. Mesoscopic structures that are associatedwith the DZ are tight-to-open folding and bou­dinage of carbonate-anhydrite laminae. As sug­gested above, the DZ is based predominantly onCastile structures, whereas in the overlying Sa­lado, deformation is nonexistent or weak; e.g.,ERDA 6.

• The chaotic seismic reflection data, which is theother basis for delineating the DZ, is character­ized by discontinuous reflectors and a blockyreturn pattern (see Sec 3.2).

The term "Disturbed Zone" has been used histori­cally in different ways during the discussion of theWIPP. The historical usage is reviewed to help readersunderstand prior discussions.

The DZ was recognized on the basis of seismicreflection data; we began with a review of proprietaryindustry data. Hundreds of line-miles of seismic re­flection data were made available in 1976 to G. J. Longand Associates (1976) for review of shallow «4000-ft)reflectors in the northern Delaware Basin. Severalseismic anomalies were identified in this review; one ofthe larger seismic anomalies was located 3 mi north ofERDA 9. At the time of the review, the extent andgeologic character of the DZ were not recognized.

Figure 1-1. Areal Extent of the Disturbed Zone, Northern Delaware Basin

10

During the latter half of 1977, Sandia NationalLaboratories in Albuquerque (SNLA), assisted by G.J. Long and Associates, made a seismic reflectionsurvey to obtain better shallow reflections and tocover the possible seismic anomalies identified earlierthrough data review. The survey, designated the 1977X Geophysical Program, covered about 48 line-miles,including some offsite work. Full details of parametersand line locations for all SNLA seismic reflection dataare contained in Hern et al (1979).

Preliminary interpretation of the 1977 X seismicreflection data revealed an area in the northern part ofthe WIPP study area where shallow reflectors werenot easily interpreted. A rough outline of an areacalled the "highly disturbed zone" was inferred on thebasis of limited data then available. The highly dis­turbed zone was reported in Powers et al (1978, Figure4.4-6). The limited amount of seismic reflection dataand lack of good reflectors prevented significant inter­pretation of the internal structure of the DZ. Theanomaly 3 mi north of ERDA 9 was included in theDZ, and was investigated by drilling borehole WIPP11 (SNLA and USGS, 1982).

In 1978 an additional small seismic reflectionsurvey ("Y" series) was conducted by G. J. Long andAssociates for SNLA (Hern et aI, 1979). A more exten­sive survey was made by Bechtel for undergrounddesign and site evaluation (Bell and Murphy andAssoc., Inc., 1979; Dobrin, 1979). These data betterdefine the boundary of the DZ and provide someindications of its internal structure.

By 1979 Sec 16, 17 of R31E, T22S were beinginterpreted as either an area of anomalous seismicdata or an area of complex geologic structure. At thattime borehole WIPP 13 bottomed at 1025 ft in theupper member of the Salado Formation. During 1980the hole was deepened to the lowermost anhydrite ofthe Castile Formation to establish the origin of theseismic signals. The hole verified that the DZ was anarea of complex geologic structures in the Castile andlowermost Salado Formations. Subsequent DZ inves­tigations have included two high-resolution seismicexperiments, the high-precision gravity survey, petro­graphic core analysis, development of tectonic models,and deepening of WIPP 12.

The seismic reflection data from 1978 are summa­rized in the WIPP Safety Analysis Report (SAR)(Department of Energy, 1980). The "zone of anoma­lous seismic reflection data" (Figures 2.7-23, 2.7-24,2.7-30 of the SAR) is used to indicate portions of theseismic records generally regarded as uninterpretableby these investigators. This zone is equivalent indefinition to the early highly deformed zone in thatuninterpretable seismic records are the bases. More

seismic data resulted in a somewhat different bound­ary to this zone.

The SAR also shows a "line indicating steepeningof dip of Castile strata to the north" (Figures 2.7-23,2.7-24,2.7-30) based on Bell and Murphy and Asso­ciates, Inc. (1979). This line approximately coincideswith the southern boundary of the DZ as defined inthis report; the Castile structure as exhibited in thenorthern part of Zone II now is included in the DZ byus (Figure 1-1). Thus, the area where the Castile/lowerSalado departs from generally parallel beds withslight dip is the DZ. For the present, we assume acommon origin for these structures. Several evaporitestructures in the Delaware Basin (Anderson and Pow­ers, 1978) may have similar origins.

NOTE: The configuration of WIPP surface control zoneshas changed as a result of the cost-reduction program, theDOE resource management policy, and Bureau of LandManagement land withdrawal actions (McGough, writtencommunication, 1983). In this report, the older configura­tion bounda~ieshave been used, in part to be consistent withpreviously published figures. Figure 1-2 is included to showthe present configuration for comparison.

III

J-----f-- ~

ZONE IV BOUNDARY

Figure 1-2. Present Configuration of Surface Control Zonesat the WIPP Site

11-12

2. General Site Geologyin Deformation of Evaporites Near the WIPP Site

by R. P. Snyder

2. 1 IntroductionThe purpose of this portion of the report is to

present stratigraphic and structural data and possibleexplanations to account for the anomalous CastileFormation structures in the northern part of theWIPP site, New Mexico. There are two main sourcesfor the subsurface data used in this chapter: (1) geo­physical logs from holes drilled by private companiesin the search for oil, gas, and potash, and (2) core andgeophysical logs from holes drilled in support of theDOE exploration program in and near the WIPP site.

2.2 Regional GeologicSetting

The WIPP site is in southeastern New Mexico,about 25 mi east of Carlsbad (Figure 2-1). It lies in thenorthern part of the Delaware Basin, an oval-shapedasymmetrical sedimentary trough almost completelybounded by the Capitan reef. The Capitan (north ofthe WIPP site) is buried, but is exposed southwest ofCarlsbad.

104°15'

32 ° 30'

32°00'

103°45' 103°15'

LOCATION MAP

NEW

MEXICO

MAPAREA

Figure 2-1. Regional Setting, Northern Delaware Basin, Southeastern New Mexico

13

2.3 Stratigraphy andLithology

The rocks of interest in this report range in agefrom Permian to Quaternary. The oldest rocks arethose assigned to the Guadalupian Series. This seriesis further defined as the Delaware Mountain Groupconsisting of (in ascending order) the Brushy Canyon,Cherry Canyon, and Bell Canyon Formations. Olderrocks are described in previous reports by Brokaw etal (1972), Oriel et al (1967), and Powers et al (1978).Figure 2-2 illustrates the stratigraphy in the WIPParea.

2.3.1 Permian (pre-Ochoan)The Delaware Mountain Group underlies the

Ochoan Series of the Permian in the Delaware Basin.The uppermost formation, the Bell Canyon, consistsof fine-grained sandstones, siltstones and minor lime­stones and shales in the basin, but near the Capitan,these grade laterally to limestone facies between thereef and the basin. (The Capitan is the reef facies ofthe Bell Canyon, and for this report is considered theboundary of the report area.)

2.3.2 Permian (Ochoan)The Ochoan Series of the Permian consists of (in

ascending order) the Castile, Salado, and Rustler For­mations and the Dewey Lake Red Beds. The lowerthree formations are primarily evaporite deposits,whereas the Dewey Lake is not.

The Castile Formation consists of interlayeredanhydrite and halite. Anderson et al (1972) dividedthe formation into seven units, from the base upward.He designated these units as Anhydrite I (AI), Halite I(HI), Anhydrite II (AII), Halite II (HII), Anhydrite III(AIII), Halite III (HIII), and Anhydrite IV (AIV). Allseven of these units are present in the eastern andsouth-central parts of the basin, but only the lowerfive (AI through AlII, HI, and HII) are present in thearea of the WIPP site.

The Salado Formation overlies the Castile. Thecontact is conformable and gradational (Jones, 1973),and is defined to be at the horizon where dominantanhydrite gives way upward to halite. In the WIPParea, this contact is generally very sharp. There arethree informal units of the Salado; from the baseupward they are the lower member, the McNutt pot­ash zone (local usage), and the upper member. The

14

formation contains several thin beds of anhydrite andpolyhalite and these beds are numbered and referredto as marker beds (Jones et aI, 1960). A thin silty unit,the Vaca Triste Sandstone (Adams, 1944) is the divid­ing unit between the upper and middle members;Marker Bed 126 (MB126) is the base of the middlemember. The middle member, the McNutt potashzone, contains minable quantities of potash; the upperand lower members contain no potash. Halite consti­tutes 85 % to 90 % of the formation in the WIPP area.Additional descriptions of the Salado and the otherformations can be found in Bachman (l980) , Jones(1973) and Powers et al (1978).

Overlying the Salado in conformable contact isthe Rustler Formation, a unit of alternating beds ofsiltstone, anhydrite, and halite. Two dolomite beds,the Culebra Dolomite Member and the Magenta Dolo­mite Member, divide the formation into recognizableunits. In ascending order, they are the lower unnamedmember (siltstone, gypsum, and anhydrite), the Cule­bra Dolomite, the Tamarisk Member (halite, anhy­drite, and traces of polyhalite), the Magenta Dolo­mite, and the Forty-Niner Member (anhydrite andsiltstone). Where the formation is buried and protect­ed from the surface and subsurface solution, the threenondolomite members contain thin to thick beds ofhalite.

Conformably overlying the Rustler are the DeweyLake Red Beds, a thick sequence of reddish-brownsiltstones. The Dewey Lake is the uppermost of theOchoan (Permian) rocks in the Delaware Basin. In thesubsurface the unit lies directly on a thick anhydriteof the Rustler. The siltstone layers show horizontallaminations and small-scale cross-laminations. Sec­ondary veins and cross-cutting dikelets of selenite(gypsum) along with greenish-gray alteration spots(reduced iron) give the formation a distinct appear­ance.

2.3.3 TriassicOverlying the Permian rocks in low-angle uncon­

formity are conglomeratic sandstones and siltstones ofthe Santa Rosa Sandstone of Triassic age. 'The SantaRosa in some places is indistinguishable from theoverlying Chinle Formation. Both of these formationscomprise the Dockum Group. Because of the generallythin layer of Triassic rock in the vicinity of the WIPPsite, it is believed that the only Triassic rocks presentare those of the Santa Rosa, the Chinle having beenremoved by erosion. In this report the term "SantaRosa" is used.

1000

4000

SOUTH:2III~..>III

Ited

Formation

Bed.

NORTH

Dne kUIIIG,ouP

"I'eftlr Anft,d,ite"llIIber of Salado

'0'1110 1/ on

Tan.II' 'orllla'ianYat.. 'o'lIIatlon

Slvln It'VIr''o'lIIatlon

Ca.t"e 'orlllation

cDD...Uoo

Bell Can,on '0'11101/ on

1000 c Cftl", Can,on '0'1110" onD..:>;;..D B,ulft, Can,on 'o'lIIatlon:>

Z "• 4000 e..• :2

Ill:

DIII

• ...·• c! D Bone Sp,ing Lillle.tone• 60004Il "~~ D0 C

0• •.. oJD4IlD

.: 8000•:a

::..• Z0

'0,000 eze>oJ>'"ZZIII

111,000 ...

,z",eViii:till.i;;;

14,0000 ~ 10 " 20

I Ii

II

II, Mil..

0 '0 20 30 Kilom.t.rs

Figure 2·2. Generalized North-South Cross Section, Southeastern New Mexico

15

2.3.4 Jur•••ic Mel CretaceousNo rocks of these age8 have been recognized or

mapped in the WIPP site vicinity. Some small out­crops of rocks of Cretaceous age have been mappedalong the Pecos River drainage system west of theWIPP site (Bachman, 1980). The Jurassic and Creta­ceous periods were evidently times of emergence, andas such, no widespread deposition occurred duringthese times.

2.3.5 TertiaryNo Tertiary rocks have been identified or mapped

near the WIPP site. About 6 mi east of the northeastcorner of the site scattered outcrops of the OgallalaFormation can be found. The Ogallala of Pliocene ageconsists of fine- to medium-grain calcareous sand­stone, and minor lenses of conglomerate. The forma­tion is capped by a resistant layer of caliche. Althoughnot recognizable at the WIPP site, the Ogallala isimportant because it is the oldest formation not struc­turally affected by movement of halite in the Castile(Jones, 1973, p 28).

2.3.6 QuaternaryThe Gatufia Formation of Quaternary age is a

flood-plain deposit and, because it was laid down on abeveled surface, some of the constituents of the bedsare from nearby Triassic and Tertiary (Ogallala)units. Pebbles from Tertiary igneous rocks of theSierra Blanca and Capitan Mountains to the north­west are present as well. The Gatufia consists ofconglomerates, sandstones, and siltstones, all poorlyconsolidated. Some massive shale layers are also pre­sent. Near the top of the formation on the east side ofNash Draw is a volcanic ash bed that was dated on thebasis of mineralogy and fission-track dating as a Pear­lette type "0" ash about 600,000 yr old (Bachman,1980). Above the Gatufia are the Mescalero Caliche, a3-13 ft (Bachman, 1980) thick unit dated at 410,000 to510,000 yr old, and windblown sand deposits.

16

2.4 Structure andThickness

2.4. 1 RegionalThe Delaware Basin is outlined by the nearly

continuous Capitan Limestone, a reef deposit thatlimited rocks of early Ochoan (Permian) age to thebasin. In the basin, the underlying Guadalupian BellCanyon Formation grades laterally into the reef rockand then into the back-reef facies (Seven Rivers,Yates, and Tansill Formations).

2.4.2 Pre-OchoanThe thrust of this report deals generally with

rocks younger than the Bell Canyon Formation, but tohelp understand the following discussion on structureof the Ochoan and younger rocks the structure con­tour map drawn on the top of the Bell Canyon isshown in Figure 2-3. Numerous drill holes have pene­trated the Castile Formation; a few others have pene­trated into the lower anhydrite (AI). Locations of, anddepths penetrated by, these drill holes are listed inTable 2-1. Geophysical logs (density, gamma-ray)were used to identify the various contacts in the oiland gas holes; core and/or geophysical logs were usedin the holes drilled for the WIPP project. Cross sec­tions (Figures 2-4, 2-5, and 2-6) show the suggestedstructure on the top of the Bell Canyon Formation inthe WIPP area as well as structure of the overlyingformations.

2.4.3 OchoanCastile Formation-The fault block shown in the

northern part of the WIPP site (Figure 2-7) is hypoth­esized on estimates of thickness of AI. The threenearest drill holes (WIPP-ll, WIPP-12, and WIPP­13) penetrated only to the upper part of AI. The uppercontact of AI in WIPP-ll is -250 ft above earlierprojections (Snyder, 1982). It is not known whetherthe block under WIPP-ll was upthrown prior to orafter deposition of AI and possibly younger units ofthe Castile Formation, because the complete sectionof AI was not penetrated.

km

miles654

6 7 8 9

Contour interval lOa 1t

234

R 35E

• Drill hole

// Fault, bar and ball on downthrawn

~ side; dashed where uncertain

o

o

Ii

IT22 S!,

R 34 E

T 21 S

R33ER32ER31E

SHELFR30E

NORTHWESTR 28E

r\·32" -L. -cL-__.,-__--=..'--__-L --..L--.:..:.N.:::EW.:..:.....:.:.M::E:;.:X:;.:IC:.:O=-- ---L------.L--a---l.....L...__..::>....._-L__-:.._--l_---r a- ...L_-L ...!- 32"

TEXAS

00

'"IT23 S

\

~0

32°15'32° 15'

0'/0

6'00

('"I

/

Figure 2-3. Structure Contour Map, Top of Bell Canyon Formation, Northern Part of Delaware Basin

17

Table 2-1. Geologic Markers in Exploratory Test Holes

[S. section; To. township. R•• range; FSl. frOl'l south line; F"Nl. from Mrth li"f'~ FEL. frOflt f"ast line; FWL. '''Of" Wf'st ltne; NP • .,ot presf'nt;leaders (---l. no l'1ata]

Hohlt LocationsNo.- S. T. R.

Geol09ic test holes IDOEI:

Coordi natesffeet I

Altitud.land sudacp

( ' ••tl

- -------- -- --------------------------------------------------_.._._.._-------------

Tot.ldepth

-- --_ ..... ----- ------------------~-------_._-----------------------------_._--

EROA-6AEC-8EROA-9NIPP-lINIPP-12NIPP-13NIPP-18NIPP-19NIPP-21NIPP-22NIPP-33NIPP-34AEC-7OOE-I

351120

9\7\720202020139

3128

21 S22 S22 S22 S22 S22 S22 S22 S22 S22 S22 S22 S21 S22 S

31 E31 EJ1 E31 EJ1 E31 E31 E31 E31 E31 E30 EJ1 E32 E31 r

2.152 rSL935 rNL267 rSL712 rNL149 rSL

2,566 rSL984 rNL

2,9A7 rSL1,451 rSL2,544 rSL1,762 rSL

202 rSL2,035 rNL

180 rSL

910 FEL1.979 FNL

177 HL29. FNL

A2 HL1.731 rEL

11 rEL13 FEL12 HL11 HL

2.430 FNL2.000 FNL2,035 FEL

60A FEL

3.540.23.531.53,40A.O3,42~.1

3.471.53,405.43,456.53.433.13.417.03.425.83.323.23,432.73,65~

3.465

NPNP1513NPNPNPNP1220

5NPNPNP

17174210o

13o

1.3925NP10o

38

721~8

51161ISS66

13806738034

154115125

538662538663628517613580560573401657662660

2.4012.9731,A247,3302,7272,960

3.0002.929

4,3062,7754,9132,8773,5AO2,7743,A481.0601,0381,0451,450

8401,8194,7344,063

Hydrol09ic test holes lODE.

H-IH-2.H-2bH-2cH-3H-4.H-4bH-4cH-5.H-5bH-5cH-6.H-6bH-6c

29 22 S 31 E29 22 S 31 E29 22 S 31 E29 22 S 31 E29 22 S 31 E

5 23 S 31 E5 23 S 31 E5 23 S 31 E

IS 22 S 31 EIS 22 S 31 EIS 22 S 31 E18 22 S 31 E18 22 S 31 E18 22 S 31 E

624 rNL722 rNL696 FNL637 FNL

2.085 FSL546 FNL498 FNL446 FNL

1,092 FNL1,038 FNL1.006 FNL

283 FNL196 rNL281 FNL

1,086 HL1.693 FNL1.660 FNL1.708 FNL

138 FEL720 FNL633 FWLtl8 FNL185 FEL236 FEL135 FEL274 FNL323 FNL374 FNL

3.397.73,377.93,377.73,377.A3.389.53,332.03,332.83.333.53.506.23,506.03,506.43.347.33.347.63,347.9

15 NP 35 502 A24-- - - -- -- - - - - - - _. _. - - - - - - - - -- - - - - - - - --Sef' H- 2c - - - -- - - - •• _. _.-- - - - - - - -- - - - - - - - ••• - .----------- -. -Seoe H-2c - -- - - - - - - - - -. -. - _14 NP 38 457 764

7 NP 20 502 A21- -- - -- -.- -. - -. -- - -- - -- -- - - -'- - - - - - -- - See P-7 - -- -" - -- •• - __ ' • _--- - - - - - - - - - - - --- -. - ••• - - - - - - - - - - - - --See P-7· _.- - -- - - -- - - - -- - _••• - - -. -- -- - - - -- - -. --. - ••••• - •• -- - --.See P-7· - --- - - - - --- -- -- - • __ • _-- - - - - - -- - •• - •• - - -- - - - - - - - -- - -- - •• -- -See P-21-- - --' --_ ... - _-- •• -- - -- -- - - - - - - - --. - -.- -" _. -. - - -- -See P-21-·- -- - -- ---- -- -- • _- •• - •••• - -. - -- --- - - - - - -- - - - - --- -- -- --S•• P-21---- -- -. -- - -- -- - _-. -- - --. - --. - •• -. - -. -. - - -- -. - - - - ••• --S•• P-13- --. -. - - - -- ---------- -- - _-. - -. --. - ••••• -- - - - - - - - -- -- - - - -- - - - - -See P-13--- -- ---- --- -- -- • • _- -. - ----- -"" - - - - - - -- - -- ---- - -- - - -- -See P·13-- ------- - ----.- • • __ • __ • _

A56563661796902415529661824925

1.076525640741

Potash resource test holes (DOE):

P-IP-2P-3P-4P-5P-6P-7P-8P-9P-IOP-lIP-12P-13P-14P-15P-16P-\7P-18P-19P-20P-21

29 22 S28 22 S20 22 S28 22 S\7 22 S30 22 S5 23 S4 23 S

33 22 S26 22 S23 22 S24 22 S18 22 S24 22 S31 22 S5 23 S4 23 S

26 22 S23 22 S14 22 S15 22 S

31 E31 E31 E31 E31 E31 E31 E31 E31 E31 E31 E3D E31 E30 E31 E31 E31 E31 E31 E31 E31 E

328 FSL121 FNL104 FSL149 FSL186 FSL

2,509 FNL514 FNL640 FNL

1.493 FSL2.341 FNL

156 FNL165 FNL110 FNL309 FSL411 FSL939 FSL

1.356 FSL139 FSL

1,652 FSL801 FSL

. 859 FNL

552 FWL171 HL

2.154 FNL1.485 FEL

160 FEL195 FNL393 FWL

92 FNL126 FEL323 FNL183 FNL198 FEL147 FWL613 FNL190 FWL

1,647 FNL398 FWL733 FEL

2,335 FNL79 FEL

130 FEL

3.345.13.479.73,3A2.73,443.A3.470.93,354.13,332.03,33A.63.411.53.509.33,503.93,373 .63.345.23,359.63,309.53,317 .93.335.93.477 .23,545.13.552.73,509.0

10IA10NPNPA

119

NPNPNPNP1210II1414NPNPNPNP

NP3HNP

A13NPNPNPII

89

NPNPNPNPNPNP

9A6A

40164

4109

146IA453966

151124

83A42323246A732

261225

35A69046A6006233573123915626A67454614273A723131638262675A780734

677I,OOA

7A6930047659630715RAI

1.0861.05R

7497216A7542646715

1.0AA1.1171,1031.043

1.5911,8951.676I.A571.8301.5731.5741.6601,7962,0091.9401.5981.5761.5451,4651.5851.6601.9982.0001,9951.915

Potash resource test holes (private industry):

0-480-1040-1200-121

0-1230-1600-\790-2020-2030-2070-2310-2330-2340-2350-2490-250A

FC-70Fe-81FC-82Fe-91FC-92

18

14 22 S24 22 S13 22 S11 22 S34 22 S36 22 S

2 23 S23 22 S26 22 S19 22 S27 22 S26 22 S26 22 S25 22 S35 22 S35 22 S

7 22 S3 22 S8 22 S

10 22 S8 22 S

3D E30 E30 E30 E31 E30 E30 E30 E30 E31 E3D E30 E3D E30 E30 E30 E

31 E31 E31 E31 E31 E

136 FSL2.5A9 FNL1.574 FNL1,'47 FI\I

2.615 FSL2,463 FSL2,655 rNL1,235 FSL2.633 FNL1,481 rSL2.365 FSL

477 FNL163 rSL

2.400 FNL2.624 FNL2,472 rNL

197 FNL166 rSL154 rSL24A FSL143 rSL

2,179 FEL1,396 FEL1.566 FNL1,301 FNL

277 FEL1,124 FNL2.655 FEL1,145 FNL1;261 FEL1,330 FNL

222 FEL27 FNL

667 HL806 FNL306 FNL

2,539 FEL

167 FEL72 FNL37 FNL

163 FNL249 HL

3,343.93,384.23.327.51.?0?6

3,431.93,305.53,243.A3,323.33,319.43,402.73,2A7.53,302.13.312.03,33A.63,265.53,2AI.9

3,382.33.470.23,3AO.53,440.43.420.6

NPtiPNP10+

NPNPNPNPNPNP

9NPNP

NPII

1StNP14NPNP

NPNPNPNP

10+NPNPNPNPNPNPNPNP

NPNP

NP12NPlOtlOt

10+5

10+SP

7010

5tlOtlOtlOt221413

1417

90133

49160115

360460360210

625240110300JOO460214260259

150292

SOl645516660635

635740655~30

987530430590610770518558563

454496

800942AI2956960

1.5241,5971.5001.?~A

1.8801,3541.3501.4431,4431,6131.3781,4061.4251.5061,3461.366

1,6031.7351.6841,7881.818

Table 2·1. (cont)

W ----------------------- - ------------~----------------------_._---------------------------------------------------

Holell locations Coordi nates A1 ti tuttf" Fom,tton top··lrtepth 1n ff"et t'lplow ,."11 surface}Ho.- S. T. R. ( footl land surfacp flatu". Santa Rosa Dewey Lake Rustler S., ado Casttl! flel1 Canyon Total

I f.otl Formation Sant1stonf" RE'd Beds Formation Formatton Format ion Formation dopth--......---------------------------------------~---._------------- -----------------------------------------------------

1-374 30 22 S 31 296 FSl 2,351 FWl 3,337.9 NP 30 305 630 1,5381-375 33 22 5 31 E 131 FNl 60 FWl 3,385.5 5t NP lOt 470 790 1,7461-376 20 22 5 31 E 118 FHl 382 FWl 3,409.5 NP lOt 40 495 830 1,7021-377 22 22 5 31 E 3,488.4 NP lOt 190 700 1,014 1,8761-383 I 23 5 30 E 2,469 FSl 226 FWl 3,271.7 5t NP lOt 160 460 1,3071-456 22 22 S 31 E 312 FSl 2,593 FWl 3,517.0 NP 9 222 736 1,070 1,9751-457 27 22 5 3\ E 195 F5l 1,270 FWl 3,458.0 NP II 146 668 988 1,8851-458 4 23 5 3\ E 2,560 FNl 426 FEl 3,378.3 HP NP 10 516 836 1,7501-459 3 23 5 31 E 2,519 FNl 1,984 FEl 3,411.4 NP 10 57 579 937 1,855

Hr-I 22 5 3\ E 207 FHl 91 FWl 3,421.1 lOt 27 140 635 935 1,747!If-2 22 5 3\ f. 651 FSl 2,069 FEl 3,403.3 875t 1,690

U-134 22 S 31 E 1,176 FSl 1,147 FWl 3,359.4 NP NP lOt 415 720 1,563

011 and gas t ..t hol .. (prhato industry):

UOC 35 21 5 31 E 660 FHl 660 FWl 3,505.0 52 516 815 2,763 4,195POGO-I 26 21 5 3\ E 1,980 FHl 1,980 FWl 3,533.0 37 548 828 2,970 4,221Me-IC 6 22 5 31 E 1,978 FNl 660 FWl 3,358.0 NP NP 9 381 691 2,509t 3,866 11,384P5lC I 22 5 3\ E 530 FSl 330 FWl 3,551.0 125 623 932 2,798 4,444W-IBU IS 22 5 31 E 2,127 F5l 2,118 FWl 3,475.3 NP 9 192 704 \,005 2,906 4,196 15,204lC-IW 23 22 S 31 E 329 F5l 330 FEl 3,585.5 NP 12 217 750 1,180 3,085 4,448 4,758G-ICOB 34 22 5 3\ E 2,022 FSl 1,978 FWl 3,453.6 HP 9 120 646 966 2,929 4,321 6,690T-\C I 22 5 30 E 974 FSl 1,976 FWL 3,329.1 NP lOt 297 605 2,477t 3,777 13,930P-IJA 2 22 5 30 E 660 FSl 2,011 FEL 3,184.1 NP NP 6t 186 496 2,256 3,610 14,900P-IJE 11 22 S 30 E 1,976 FNL 1,981 FEL 3,206.1 NP 17t 254 577 2,310 3,642 13,8525-IJR 36 22 5 30 E 660 F5L 2,006 FEL 3,305.1 NP >31 199 506 2,386 3,816 16,252C-15 2 23 5 31 E 660 FHL 661 FEL 3,443.2 NP 5 55 615 1,060 3,018 4,419 5,191G-ICAB 5 23 S 31 E 2,062 FHL 2,089 FEL 3,328.1 10 NP 18 327 657 2,590 4,045 4,150B-4JR 6 23 5 31 E 2,180 F5L 330 FWL 3,296,0 NP 192 522 3,882 14,354C-7JR 6 23 5 31 E 1,957 FHL 1,973 FEL 3,319.4 NP 26 286 606 2,576 3,951 14 ,550B-IH I 23 5 30 E 1,834 FNL 1,978 FWL 3,284.7 5t NP 1St 148 465 3,777 14,292B-lJR I 23 S 30 E 1,973 FSL 1,648 FEL 3,288.8 5t NP 15 172 482 3,871 15,3758-IOJR I 23 5 30 E 1,980 FHL 660 FEL 3,299.0 5t HP 11 184 517 2,479 3,824 14,306

JiKoy to abbreviations:

$"lurces of well-log data

OOE OopartJnont of Enorgy PSLC Pogo Producing Co., SlC Federal '1AEC At...!c Enorgy C..... ission W-1BU Clayton W. Will iams, Jr., I-Badger Unit rederalERDA Energy Research Ind Development Administration TC-1W To..s Crudo Oil Co., 1-23 Wrf9htWIPP W.sto Isolation Pil ot Plant f.-lCOB Michael P. Grace. I-Grace Cotton Baby FederalH Hydrology T-1C Troporo 011 and Gas Co., I-Cabanap Potash P-1JA Phillips Petroleum Co .• I-James "A" St. to0 Ouv.I Corp. P·IJC Phfllfps Potroleum Co., I-J....s "C"rc Farm Ch... i cal Resource Development Corp. 5-1JR Shell 011 Co., I-Stato J.mos R.nchI Internatton,' Mtneral s and Chemical Corp. C-1S Continont.l Oil Co., I-Stato, AA-2HF National Fanners Unton Service Corp. f.-ICAB Mtchael P. Grace. I-Cabin Baby FederalU Uni tod St.t.. Potash Co. B·4JR 6elco Petroleum Corp •• 4 James Ranch UnitPOGO-I Pogo Producing Co .• Federa' '1 C-7JR Continental 011 Co •• 7 James Ranch UnitUOC Union 011 of C.'ifornla, Fedor.' Fl '1 B-1H Belco Petroleum Corp .• 1 Hudson FederalMC-IC 8ryon McKnight, I-Campana Federal R-3JR Relco Petroleum Corp .• 3 James Ranch Unit

R-I0JR Belco Petroleum Corp., 10 James Ranch Unit

19

I} Phil/ip. Pet. Co., Sandy Unit # 12} Be/co Pet. Corp., Jam.. Ranch # 'a3} Shell Oil Co., Jam.. Ranch Unit # 14} WIPP 135} WIPP "6} ERDA 6

7} Union Oil of California, Federal FI # 18} Pogo Producing Co., Federal # 19} Fred Turner Jr., AID Federal # 1

'O} Phi//ip. Pet. Co., James "c" # ,ff} Bryan McKnight, Campana #' I12} Clayton W. William. Jr., Badg., Unit F.deral #1'3}Texas Crude Oil Co., Wright-Federal # '-2314) Ralph Low., Bass F.d.ral # 1

10

B

7

9A'

8

6

'eI -'...4 B I

WI P P 13

SIT E

I2

2 mil ..I

20

Figure 2·4. Index Map Showing Lines and Drill Holes Used in Cross Sections Shown on Figures 2-5 and 2-6

EXPLANATION

Permian(Ochaan)

}Permian(Guada'upian)

} Quaternary

1"''''''

Castile Formation

A= Anhydrite

H = Holite

Sonto Roso Sandstone

Alluvium, caliche, sand dunes,and Gatuna Formation

Salado FormationPst- Fletcher Anhydrite in

backreefTronsi tion zane -interbedded halite

and anhydrite ot bose at unit

Chinle Formation

Rustler Formation

·:H I:.A.I·

.:Am:.'or

:A U·

~ Dewey Lake Red Beds

Pbc - Bell Canyon FormationPel - Capitan Limestane(reef)Pt - Tansill Formation

l----J'--__--l Py - Yotes Formation

fi-~-------

'==:p:5===Psf

EJA'North

9e

Bend inSection

7

Bend inSection

6

Bend inSection

5

Bend inSection

Rsr

4

Bend inSectionBend in

Section

2

Bend inSection

-:-:-:-:-:-:-:-:-:-:-:-:-=-:-:-:-:-:-:-:-:-:-:------------Pr

i-----__---------~t_-~r---------_::Pdl

2500

2000

1000

3000

11-11§§§iIIi~~~~EIIl@

::::::::::=:::=:=:::::::::::=-:=:=:=:=:=:=::.":::=:=:=:::::=: :::=:=:=:= =- :-:-:-:-:--=-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-=-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:=::-:-:-:-:-:-:-:-:-:-:-=-----=------:-:-:---------~ - - - - - - - ------~------- - - - - - ------------~-- - ------~--~-~--

1500 ~~~~~~~~~~~t~~~~t~~~~~~~~~~Itr~~=~~~~1~~~~~~~~=~ =t~=~~~~~~~~~~~~~~~~t~~~~~~~~~~~~~~~~~~=t~~~~~~=~~~~tII~~~~t~~~~=IIIIIff~ft~~~I~~~~IIIII~~~~=~~~~~~~=t~~:~~==~~~=~==~=~~~- ~~~~~=~=~~=~~~~

---=----=----=-~---=----=----=-:---=----=----=------------------------~----------------.~=======::::=============:c:=======::::======== -- ....>.;--:-::=-:==:==:==::::~::=:::=:=-:::=:=- ..

. .. .. . ... .. - - - - - - - -----------------------------------------~--~--:--.,.-~---

3500

ELEVATION

ASouth

300

800

600

200

500

900

400

700

500

100

m.t'rs te.t0 0 300 1000

Pel

-100 200

-500 500

-200 100

Pbc ? Pbc 2 mil.s-300 -1000 Pbc0

kilomet.rs2 3

? ?-400 ?

-1500

meters feet

1000

Figure 2-5. North-South Cross Section, WIPP Site

21

8Weat

8'Eaat

'00zoo

L_-r_.L..-__,..J2 milll

S IF.llolll.'.,.

'00

(s•• flgur. 2.4-3 for ."planatlonof aymbols'

met.,. fu.500 1000

141312

Se"d ,..S.ction

"IS -10 ....

.lO~

---~ POI

P,

III,:::::: % "",,=:~::S=:±=:~

~I '\.

0i

:± I.2ill-----2

?1(~ 2ill2?~? Pbe

S...dlnS.ctlonEUVAT'ON_'.r. f ••'

.100

1000

000.000

100

1000

700

0002000

300

1500

400

.00 1000

200

.00

.00

0

-'00

-300

-200

-300 -1000

-400

-'300

Figure 2-6. East-West Cross Section, WIPP Site

22

'1

o 5000 10,000Lol...._ ...' .....,........'.-..'.........'.-..'...JI FEE T

o

I

D

til ,

t c:>

.~ I!J ~ (,) C!l

o

co

R3.1 Et

I

G~l 0

II1---­I

II------------ .-,---

I

I

II

.+-- .. -, <>-

III

I :::':: ~·ou;~t::~~:nd.., "d., QUlrl.d .It."

t I uncI".I".

o Drill Ito"

-<?- Dr, Ito"

(} Producing " •• ••11

+

(contour Int.rrel '00 til

Figure 2-7. Structure Contour Map, Top of Anhydrite Unit AI, Castile Formation

23

The northwest-trending fault at the southwestcorner of the WIPP site is an interpretation to explainthe abnormal thinning of AI in this area. In one drillhole (B10-JR) the total AI unit penetrated measuredonly 185 ft thick. Holes nearby have AI thicknesses of253 ft (S-lJR) and 237 ft (G-1CAB). The majority ofthe structures in and around the WIPP site appear tohave been the result of movement of halite (HI andHII) in the Castile.

The structure contour map drawn on the top of AI(Figure 2-7) mirrors the structure on the Bell CanyonFormation.

The structure contour map drawn on the top of HI(Figure 2-8) shows the beginning of the chaotic distri­bution of the Castile halites. A broad area at theWIPP-13 drill site appears to be depressed 400 ft.Halite I is 94 ft thick in WIPP-13; to the southeastabout half a mile, HI is 516 ft thick in WIPP-12. Itappears that there has been movement of HI from theWIPP-13 area to the WIPP-12 area.

Northeast of the WIPP site at the ERDA-6 area,the structure on the top of HI shows a northwest­trending anticline. It appears that halite from HI hasmoved upward stratigraphically > 1000 ft.

The isopach map of HI (Figure 2-9) shows athinning of HI at the WIPP-13 and WIPP-ll loca­tions, a slight thickening at WIPP-12, and a majorthickening at ERDA-6.

Overlying HI is All, a generally 110 -120-ft-thickanhydrite layer. The structure contour map drawn onthe top of All (Figure 2-10) is similar to that of HIbecause this thin anhydrite usually was carried alongwith the movements of HI and mimics the HI struc­ture. The surface of Halite II (Figure 2-11) overlyingAll, also shows a low in the area of WIPP-13 and ahigh at ERDA-6. Additionally, there is an oval high inthe WIPP-ll area. It appears that halite from HII hasmoved from the areas of WIPP-13 and ERDA-6 to­ward WIPP-11. The cross section (Figure 2-4) showsthe interpretation of this. The isopach map (Figure2-12) of HII shows that the majority of the halite thatmoved toward WIPP-11 came from the area of WIPP­13 and from the area between WIPP-ll and ERDA-6.

The structure on the top of the Castile Formationis shown in Figure 2-13. The three major areas, highsat WIPP-ll and ERDA-6 and a low at WIPP-13, areapparent although there is some muting of the struc­tures. The AlII isopach map (Figure 2-14) shows someinteresting features. The thickness of AlII at WIPP­13 is excessive. Core from the drill hole shows dips of45° or more on the laminations. The cross section(Figure 2-4) depicts the low point of the base of All tobe located at WIPP-13. This is probably not the case;

24

the low point of the base may be offset to one side orthe other. Obviously, the cored thickness is not thetrue depositional thickness; the true thickness of AlIIat WIPP-13 is probably 350 to 400 ft.

At WIPP-ll, AlII is -200 ft thinner than normal(Figure 2-14). This may be due in part to stretching ofthe anhydrite as HII rose underneath and in part tonondeposition as HII may have been rising prior todeposition of AlII. In the ERDA-6 area the structurehas been interpreted as a diapir. AlII has beenbreached by the upwelling of the underlying HI (Fig­ure 2-4). This interpretation of stretching and conse­quent thinning of All had previously been made byAnderson and Powers (1978).

Salado Formation-The Salado Formation con­tains many thin beds of anhydrite and polyhalite,some of which have been used in this report in theconstruction of structure and isopach maps. A 30-ft­thick anhydrite bed, the Cowden Anhydrite, in thelower unnamed member of the Salado, is one of thesebeds. Other markers used are the base of Marker Bed126 and the top of the Vaca Triste Sandstone. Theselast two units divide the Salado into three units: theunnamed lower member, the McNutt potash zone(local usage), and the unnamed upper member.

The structure contour map drawn on the base ofthe Cowden Anhydrite (Figure 2-15) shows a remark­ably smooth surface when compared with the previousstructure maps of the various units in the CastileFormation. There is a low at the WIPP-13 area, notmuch structure at the WIPP-ll area, and a broadsoutheast-plunging nose at ERDA-6. Southwest of theWIPP site there is a small oblong high.

The isopach of the interval between the base ofthe Cowden and the top of the Castile (Figure 2-16)when compared with the structure contour map of thetop of the Castile (Figure 2-13) looks as if the intervalwas deposited on already-existing and still-developingstructures and deposition of the interval tended tolevel out the hundreds of feet of structure on theCastile Formation.

The structure of the top of the lower unnamedmember of the Salado (Figure 2-17) bears almost noresemblance to the structure maps of the underlyingunits. There is a very slight high at the WIPP-ll siteand an oblong northwest-trending high at ERDA-6.The small closed high off the southwest corner of theWIPP site is still evident. Most of the structural highsand lows seen on underlying surfaces have disap­peared. Instead, a very broad northeast-plunging syn­clinal surface has developed in the eastern half of thesite.

R31Et

~¢

#) ) 0 C

",00

r:-'oo~

¢ (t0-"""t - -P...,

WIPP-I/ L(t

,L~

(;) ~, - 200

F..J d L....,I

I " L "'4"»

+ ! , "»--- .0 + CIl

I

rJ0 I

a I 0

~I

+0 . ;1-+--- ---II 0

1.0 00 i

~~ I01 '

~L oj-<>-

0 0 1 r .CII IL r J

r _l--Ja 0 1°()

r JGl .~

I 0 Drill hoI.,

~ <:> 0 0 -<r Dry hoI.I

~-() ProducIng g•• .,."

SAND 0--- -e-~ 0 0

.,0 5000 10,000I , , , 1 I , ,

t 1 FEET

(contour Int.rrel 100ft; 'liechur•• Indlcet. clo••d lorr)

Figure 2-8. Structure Contour Map, Top of Halite Unit HI, Castile Formation

25

R3'Er

'?0

rJ

((~--../OO

.,0

I

~- ......

r

0 I ............

L400

o ~J/l--, ""o r 200------ - ~OOr.J G G D • L...,~ Lf5' 300 1 -, 0"" T

I >-1>-...,0I c <r L ~.~ -<

~I~o 1 1 84:' OJ

J r I 6refr 0 I ~~ yI '-.L.J-..Y "'4

t0 f~' ., \0 j ~ ~ + ~

I ~ ellI q I:

01 v tJ -e.·J ,-,1 +L, J ¢

o 0 r IQ 0 IL '" I-0-' )' r J I

f I 0 1 Io ':ot- --, ., I ~ _.-.J

o r °1 0o ...,0 I

~ r JQ I / L .......1 ,

\

0 Drill hole

l 0 """- 0 X 0 -¢- Dry Irole

¢: Producing g•••ell

r b SAND \. 0 ¢> •el !l f) (~ ·t

o 5000 10,000I • I • I I . I I I I FEET

(contour 'nt.r".', 'OOft, h.chur•• Indlc.t. c/o••d lowl

Figure 2-9. Isopach Map, Halite Unit HI, Castile Formation

..,26

o

r..JI

'-' l

r J0

u I

r !-II

I) ,

10 0

L1(:J

La

I)0

Q

Q

~ 0 ()

r l!l I!> ,)

R31E

---'~

SAND

o 5000 10,000'L...-a'........-Io......, ....L.1~...........! ......! ...JI FE E T

°

:/1/0~o I, I

o Drill hole

-9- Dry hole

-0 ProducIng gas we"

(contour Int.",e/, 100ft; hechur•• Indlcat. c/o••d low)

Figure 2-10. Structure Contour Map, Top of Anhydrite Unit All, Castile Formation

27

o

R31Er

o

<)- i

t

I

~iI

I+

II+I

~o ¢> .\00 I).-/ ~

c

SAND

o 5000 10,0001..1 ...l'--I.'......' ...l'i-...J.'-L.'................' oJ! FEET

o

o

oI

o Drill hoI.

l -¢- Dry hoI.

(:f Prolluclnll II" w."

(contour Int.r".1 100ft)

Figure 2·11. Structure Contour Map, Top of Halite Unit HII, Castile Formation

28

R31E

- 0,

c

~;

tI

I,+

o Drill 110/.

-<?- Dry 110/.

a Produc/rtg ga • •• /1

c

"00 I

~«0 Io '

II

,_,..J

o

rI

r J~

c

(;

- --,0,L..,

L..- -_.. -- --

0

0

f5I (

'-' l

r J0

I

~j -

II

'" I0

L10 Q

L_O'0 ~

00

Q

3l (' f)

/0

r....

oI , ,

5000 10,000, I , , , ,I FEET

(contou, Int.",al, 100ft, hachu,•• Indlcat. clo••d low)

Figure 2-12. Isopach Map, Halite Unit HII, Castile Formation

29

R31E

I~

¢.t

I

I

I

+¢

j

III

Drill hole

Producl"g g•• ",ell

I6'00 -,'----

~~ II .

o-<?- Dry hole

o

r "oo II'------

. + ¢II!

!

. .

l_¢__ ~SO-.oo~

':>00 ')

~

o 0

o 5000 10,000I " ,I, "I FEET

(contour Int.r.,a/, 100ft; hachur•• Indlcat. c/o••d low)

Figure 2·13, Structure Contour Map, Anhydrite Unit AlII, Top of Castile Formation

30

R31E

"1">-->--2'Z

616 (> III<..l'<..l t:;:,.1cto'w ...~I--'

+ ~i

+I

II

+ ¢

IIII

..Produclnll II....ell

o Drill hole

Dry hole

I

L,II

r

_ • ..,J

1" ··0,

,WIPP-12

~ ~!lOo

8'

11

SAND

~~.

G

0

rr J

p.J

I l

0 l

rJ0

I!-

o

o

° 5000 '0,0001.1 J.. ---l. FEE T

(contour 'nt.r".', 'OOft; h.chur•• Ind/c.t. c'o••d 'ow)

Figure 2-14. Isopach Map, Anhydrite Unit AlII, Castile Formation

31

R3'E

¢-!t

I .I~ ,

00 I

I+---1°° !I

+II __' ......

..... /1

1II

o

o

/loo~r.:9 .

----~I

100 '0---4----

I

r

r

,/

,/

//

I/

/ .,/.,0

10q I

/ifll

\\\

'200~

<> ~O .....I /6

/ /.~ / I__.-.L_ / ' '

/ / '600~j

/ /~! I/ / I/sAND / 0 ¢> i

/ /.)...

0L..' 5_0..,O_O '_0..J'?OO FEET

\"- ...... -

o

o

o

I-0

1° 1

/ L,{o L-

~-------, : /o Drill 1101. 7

-<?- Dr, 1101.

{) Producl"g g•• ••"

(f;oluour 'nt.rv.', tOOft, h.chur•• Indlc.t. c/o••d low)

Figure 2·15. Structure Contour Map, Base of Cowden Anhydrite Member, Salado Formation

32

l._.._

R31E

o

o 0,,,, Ito/.

-<?- 0', Ito/.

() ",odue/nll II" ••"

<>.+

150 _ -+- - - -t--- .--o¢'- I·t .....-. ....

o

r- _......J

1r j

.---l

- 0

° 5000 10,000�1.- .....' ~I FEE T

o

o

o

L,Q

-t- -/50 _ L_........ ()

(contour Int.r ••1 100ft)

Figure 2·16. Isopach Map, Base of Cowden Anhydrite Member to Top of Castile Formation

33

Produclnll 11•• ..."

o Drill hoi.

-<?- Dry hoi.

ao¢­./

i-R3'E

\)

,I

---J-/

//

/SAJD

o '000 10,000....' ....' ....1 FEET

°1 Il..,1

'f----/

//

,) /./

;

io

- - ~ - -- I,pO.,- "ERDA-6 0

.// -- ~- ...................~): ....... \.r-------r'--------¥- '- I \

,",- __ / 1800 '- - - - - ./"'- .,---........ .,-~" I

( .,-- '-,/ " 1800~I ~ ..........-....,...:;,...,..-,.-+-"""1r.---~-, ,/-, :9---t- .........

J .// r" L I .'J ......._~OO~'..................

/ / 00 Lf1 1'-/ r / ,1 /, / 1-.,

1600K'""

/ .{I.J / ( ~"~ ~ ~ / ' i "" ~r " ,./ ,

./ I" ' - >-k:/ I· '\ <t L,/ ~z .(\ 1 / 616'- II)o ",,., ..;WIPP-IZ f' r U'u'- 0, C\j

1 \ '" \ !"".-- O~>-T~--- -..:. ~I r '. ./ ,:>0 .w

\ I/'I ' W( I ~ / '\ I I) / / '.' /-t A"

'i~" \" V / / /' {', Vi-I 1 \ / I /'

, \ I \ I 1"00 -I 0,

,., I "" \ \ \ I'\ \ r \ \ I (,~

\ \ \ \ \\ I \ /0 \

r~~~~1r"J,:-1-+:-. ..-......;.I-+o----\~___,I__Q-...!..I-.,......J.?_--LI _-+__"~r +.- .....J

I' \j \

)./

,//

//

Figure 2-17. Structure Contour Map, Top of Lower Unnamed Unit, Salado Formation

34

The isopach of the lower unnamed member (Fig­ure 2-18) makes it appear as if depositional filling ofexisting structural lows and depositional thinningover existing structural highs of the Castile (Figure2-13) was responsible for the thickness changes in thelower member. Structure on the McNutt potash-zonesurface (Figure 2-19) reflects only in a vague mannerthe underlying surfaces (note the decrease in contourinterval needed because of decreased vertical struc­ture despite the increased number of control points).The structures at WIPP-ll and -13 have disappeared,the plunging nose at ERDA-6 is present but muted,and a minor high is present offthe southwest corner ofthe site. The isopach map (Figure 2-20) shows a rangeof 100 ft of thickness over the WIPP site.

The structure of the top of the Salado Formation(Figure 2-21) looks much like that on the top of theMcNutt. A broad southeast-dipping synclinal struc­ture trends across the eastern half of the WIPP site.This structure begins to appear on the lower memberof the Salado and becomes more pronounced upward.Neither the isopach map of the upper member of theSalado (Figure 2-22) nor the isopach maps lower in theSalado show a thickening of the formation over thissynclinal feature. The rapid thinning of the upperSalado northwest of the WIPP site is caused by disso­lution of halite at the upper contact of the formation.The line showing the easternmost extent of this disso­lution is shown on Figure 2-22.

The numerous marker beds of anhydrite or poly­halite in the Salado Formation can be used for aninteresting comparison. Thicknesses of halite betweenthe marker beds between drill holes in the WIPP areashow some apparent patterns for 20 of the drill holes,as shown on Figure 2-23. Most of the 20 drill holeshave one to three intervals that are thinner or thickerthan normal. This is understandable in terms of depo­sition of halite. Two drill holes, WIPP-ll and B-IOJR,are anomalous in that WIPP-ll has thinner halite for18 intervals, nearly all of which are in the lowerunnamed member, and B-IOJR has thicker than nor­mal halite intervals in seven of the same intervals. Indrill hole C-7JR, about 1 mi east of B-IOJR, five of theseven thickest intervals are in the lower Salado. Theexcess thickness of lower Salado units in the two JRholes may indicate that some causal relationship ex­ists.

Because it is improbable that the thin markerbeds in WIPP-ll were deposited on highly dippingsurfaces or shallower underwater surfaces, only tworeasons can be postulated for the abnormal thickeningand thinning. In the case of WIPP-ll, where there is a

structural high on the Castile, the basin surface couldhave been slowly rising during early Salado time, orbecause of the suggested post-Triassic, pre-Pliocenedeformation, halite could have been squeezed out ofthe halite intervals, leaving a thinner than normalsection. Either case would explain the abnormal thin­ning. In the B-I0JR area, the thicker lower Saladobeds may have been deposited in a slowly down­dropping surface. The area at C-7JR may have beenaffected the same way slightly later in lower Saladotime.

The Castile structural high at the ERDA-6 areadoes not appear to have affected the lower Salado bythinning the halites. The "Infra-Cowden" is thinnerthan normal here, but only because the diapiric struc­ture of the Castile actually moved upward into theSalado, forcing some "Infra-Cowden" halite aside. Thearching of the formations overlying the Castile with nothinning of those formations indicates that at leastsome of the upward movement of the Castile halitesoccurred after deposition ofthese younger formations.

There is no similar arching over the WIPP-llarea, and either there was no post-lower Salado move­ment or, if movement occurred, the halites in thelower Salado were squeezed outward and therebyabsorbed the vertical movement that was not trans­mitted upward into younger formations.

Rustler Formation-The surface of the RustlerFormation (Figure 2-24) appears much like that onthe Salado. The oblong high at ERDA-6 is present butnot prominent, as is the high off the southwest cornerof the WIPP site. The southwest-plunging syncline onthe Salado in the eastern half of the site has become aclosed low. The southeastern part of the syncline isnow higher than the northeastern part. This can beexplained by the thicker Rustler in the southeast partof the site. The isopach map (Figure 2-25) of theRustler shows the greater thickness in the southeast.Dissolution of the halite beds in the Rustler is pro­gressing from west to east across the site. As the haliteis removed, the formation loses thickness. On thewestern half of the WIPP site, all of the halite hasbeen removed and the anhydrite beds are hydrating togypsum. This adds thickness to the· formation. Tocomplicate the isopach map even further, some of thenewly created gypsum is being dissolved. This three­step alteration process causes the "hummocky" ap­pearance in the western part of the site. Figure 2-25shows the areas of the WIPP site where these dissolu­tion stages occur. The closed contour at ERDA-6 isalso a reflection of dissolution of halite rather thandeposition thinning over a preexisting high.

35

o '000 10,000""I ~I ---II FEET

o1000 .

. 54 No-.....,.. .

~:t

CIlC\I

V C\It ...

I

~II

o Drill hoI.

-<:r Dry hoI.

e ~ () ProducIng g•• .,."

U~--_-L-_-_---J

o

rI

r J.-.J

.>

o

- -:-lf)1

L.,L.--~_. __

Q

o

o

o

o

0 (;

iii

) 0

'" l!J ~ 0

Figure 2.18. Isopach Map, Lower Unit, Salado Formation

..

36

o

G)

~\'II

"" 0\'IIIJI

t< \ l!>I

-R31E

\\

\.--- .......... J

o 5000 10,000L,.! .L.' ---', FEE T

"'f- Or, hoi.

:0 Producing g •• ."."

(contour Int.r"e/, 26ft; hechur•• Indlcet. c/o••d low)

Figure 2-19. Structure Contour Map, Top of McNutt Potash Zone, Salado Formation

37

R3'E

<)+

~: ,

. '" vI ~.,.

III

II

+-

'''0o

o Drill hoi.

-<r Dry hoi.

o Producing g •• ."."

o

,) .... I~z'z

..; w,pp·,zgig

<) CIlu'uf' ~ >-lct CII

f) CII0, ... to-~I--'

SAND

_____-400

- ---,°1

l..,L-.-- _

rrs.J

I (I

r J

I! -

o

o

'J

~1~

i 0

0

0..., -..,.__5_0...?_0 '_O....,?00 FEE T

(contour Int.",el, 26ft; hechur•• Indlce" clo•• d lorr)

Figure 2·20. Isopach Map, McNutt Potash Zone, Salado Formation

38

I

r

R3fE

° ~OOO 10,000...' ---'"' ---'1 FE E T

{contour Int.r".',26ftj h.chur•• Ind/c.t. c/o••d lorr}

I/)l\jl\j...::r------r

-<?- Dry hoI.

{) ProducIng g•• .,./1

Figure 2-21. Structure Contour Map, Top of Salado Formation

39

:'\.~

o

r

o

--,L-.,

1....,L..,,L

II

iII

u J

rJI

r J

_ • ....JrI

r J---l

1"

..., l.-

!lOO

.. O.

r} WIPP- 12

o :5000 10,000L.!__---1.'---.....11 FEET

j

I

I

I1--+-+-~--.-ll,,-----~-.>.l.-::~- ._~ .--..--.-------~-+_----'--_+_---+

I

II

~

I

III

>-1>-......

o

o

o

I U

f

o Drill 1101.

~ Or,IIol.

i) ProducIng g•• ..."

Figure 2-22. Isopach Map, Upper Unit, Salado Formation

40

R3fE

Q

r

o

<>POGO-'

~UOC

ERDA-S

IIIII

I

I

AEC-7

I+

I

II

+III

I

II

II

II~.

I

~I~1-1-z·z6/5u·ug..;0 .....

81--'

Drill Ito/.o-<?- Dr, Ito/.

a Produc/n" " •• ••11

.:<> P-SCL

o

.J

·t

o

W-18U

<>-

D

rI'

r J_ .......J

DOE -1c. ~. .J

W/PP·12'-l<)

0.

q

("1 WIPP-!J

o 5000 10,000...' .....&.' ......' FEET

'0

I .'

o

SAND

o

- ~C-7JR <>

l.., "'L-- _

<} r-IC

o

Q

o

QP·'JE

o

P·IJAo

o

,

~ 0

o

0 (,

iii

) 0 0

r l!) I!> f)(0)

Figure 2-23. Location Map of 20 Drill Holes Used in Study Of Salado Formation Marker Bed Intervals (0- Wipp-ll with 18thinner than normal intervals,A- B-lOJR and C-7JR with 7 thicker than normal intervals)

41

Figure 2·24.

42

o

Product"" "•• ••11

o. ' !SOPO '0,000

(••",our 'If' ' FffT.r".'. 2'''· h• .chur•• ' dSt,."", Conto", M • I•••••, .... Ie.)

ap, Top of Rustle F .r ormatIOn

"31Er

r

;1

;)o

a

"a HALIT£

b./oll

G CUL£.RA

I""", "

q

"

~ t- --+325~

II~

I

o o

(>

SAND . 325~

o 5000 10,000I I I FEET

Dr, holt

Producln/l /la...a/l

(con,ou, 'n,.n.', 2'''; It.cltu,•• Indlc.'. clo••d low)

Figure 2·25. Isopach Map, Rustler Formation

43

Dewey Lake Red Beda-The Dewey Lake RedBeds are the oldest unit where the structure (Figure2-26) and thickness (Figure 2-27) are more a reflectionof surficial erosion than of underlying structure. Fig­ure 2-27 shows the influence of surficial erosion on thewestern half of the WIPP site where the overlyingSanta Rosa Sandstone has been stripped off, thusallowing erosion to occur. On the eastern half of thesite, the surface of the Dewey Lake reflects some of thestructure of the underlying Rustler Formation.

2.4.4 TertiarySanta Rosa S8ndstone-The structure map on

the Santa Rosa Sandstone (Figure 2-28) shows little inthe way of underlying structure. The formation hasbeen eroded from the western half of the site. Thethickness map of the unit (Figure 2-29) indicates thatthe underlying synclinal feature in the eastern part ofthe site may have been filled in by thicker deposits ofSanta Rosa material. The southwest-plunging troughin the southeast corner of the WIPP site may reflect adrainage channel that developed on the Santa Rosaduring or after deposition.

2.4.5 QuaternaryGatufta Formation-Outcrops and deposits of the

Gatufia Formation are sparse in the WIPP site area. Astructure map shows very little that can be used in theinterpretation, and no map is included here. Theisopach map (Figure 2-30) is included here only toshow the sporadic depositional pattern of the forma­tion. The map Sh3WS in a general way that the forma­tion was deposited in topographic lows on the SantaRosa and Dewey Lake. Erosion of the Santa Rosa hadprogressed to the point that it had been removed fromhalf of the WIPP site prior to Gatufia time.

2.5 Discussion andConclusions

Definition of the structure of the rocks under theWIPP site is the main purpose of this chapter. Thisdefinition includes both the actual configuration ofvarious units and, if possible, a reasonable time frameduring which deformation of some of the units tookplace.

The major structural deformation, as presentlymapped, occurs in the Castile Formation. The two

44

cross sections (Figures 2-5 and 2-6) illustrate thedifferences in thickness and elevation of the variousmappable units of the Castile. The structure andisopach maps of these units show the same featuresfrom a plan view.

It is apparent that the major cause of the struc­tures under and around the northern part of theWIPP site has been the flowage of the two halite units(HI and HII) of the Castile. Movement on the inter­preted faults cutting the lowest anhydrite (AI) mayhave been the triggering mechanism for this flowage.Another possible triggering mechanism may havebeen the eastward tilting of the Delaware Basin duringmid- to late-Cenozoic time (Powers et aI, 1978). Over­burden, possibly including Cretaceous rocks not nowpresent, may have helped in the Castile halite move­ment by adding sufficient weight in places over thehalite to initiate movement (Jones, 1973). Jones(1973) has bracketed the age of Castile deformation toa period from post-late Triassic to late Cenozoic (pre­Pliocene). There seems to be general agreement thatmovement of the halites of the Castile occurred some­time between the time after deposition (late Permian)of the Castile and before the time of deposition of thePliocene Ogallala Formation.

Study of drill and geophysical logs shows that inthe lower Salado, below the Cowden Anhydrite, thereis a general thickening of the section in areas wherethe underlying Castile is structurally lower than nor­mal. The geophysical logs also indicate numerous thinbeds of anhydrite below the Cowden Anhydrite inthese areas. It appears as if the structures were form­ing as the lower Salado was being deposited, and inminibasins over the Castile structural lows, some an­hydrite was being deposited along with thicker halitebeds. The cross section (Figure 2-4) shows these de­posits as "transition zones" at three drill holes,B-JRI0, WIPP-13, and POGO-I. WIPP-13 has core inthis "transition zone," for the other two density logswere used to identify lithology. This lower part of thelower member of the Salado, called the "Infra­Cowden," is one of the most difficult lithologic units tointerpret. Southward from the WIPP site the CowdenAnhydrite appears to cut across the lower Salado andmerge with the uppermost Castile anhydrites (Jones,1973, p 15). This relationship is shown on Figure 2-5on the south side of the cross section at the SandyUnit No.1 drill hole. Across the WIPP site the Infra­Cowden thickens and thins as the underlying Castilesurface lowers and rises. This change in thickness ofthe Infra-Cowden gives the appearance of a unit thatwas deposited on an uneven surface of the Castile.

o

,0 <:>

~ l!l IIfI 0

R3'Er,

W •• t.rn .dg. ofSonto Ro.o

SondstonelJ

- --,°1

L,L.------

o

SAND

° 5000I I

+

1

>1> II)

........ CIIz·z CII

515 4...

U'ub

;=:t3315<r

II

I

l I

rot .--+---.---. <rI

I Ir J

I4 I

II

0 Drl/l hoI.

-<f- Dr, hoI.

0¢ a ProducIng g ....."

"10,000I FEET

(contour Int.nal, tlI,t; hachur•• Indlcat. clo••d low)

Figure 2-26. Structure Contour Map, Top of Dewey Lake Red Beds

45

Producing g•• ",.11

(contou, Int.,,,.1 25ft)

l"7~............. ~----------.

0~ 0-t -«- Dry hoi.° ~OOO 10,000

L..I ......l.' .....J' FEET {)'---- .....J

r-.,

Figure 2·27. Isopach Map, Dewey Lake Red Beds

46

R31E

0 0,'11 1101.

-<?- 0" Itol.l

a ",oducl"g g•• ..."

I-- ~ P> 'J 1(.,

')

I

0­t

~jII

~I~1-1-z'z616l)0l)

~c,..,:;::1-'

ItI

I

I

t<>

j,II

II

+--II

I

<>

o¢-

vERDA-6

SAND

Erosiono/ edge ofSonto Roso Sandston

\\II

a 5000 /0,000...' - __---I.' .....J' FEET

r)

--,0,L,L- _

o

o

rr.J

r..JI (l

r J

I! .II

I') IL1

~

L-o.,

o

o

o

u

o

io 0

(contour I"terrel 26ft)

Figure 2-28. Structure Contour Map, Top of Santa Rosa Sandstone

47

r-·.....J1("

r--',.--J

//

/I

We.terll edge of

SOlita RosaSOlld.tone

o

R31Er

o 5000I I

SAND

o

- -:-li)1 <)

l, (')L----- __

o

o

o

oo

o Drill hoi.

-<}

0(}

r-a-i) r-....J

oJ

0 rr.J

p..J

I (;

0 1 -Q-' II).t -.-<\1

rJ <\Il0-

a\

0 I 0 'J

+" ~tI 0

I01 0 \J~l

0 0EI

L_i)'

it

-<r Dr, hoi.

{$} Producing gu ... /1

. .~ I'>

(contou, 'nt.,,,., 26ft)

Figure 2-29. Isopach Map, Santa Rosa Sandstone

48

+RI1E

r I• I

0 I,>

0 f0 0 , ~

~I 0

~RDA-6

I

I

I

II

<>(t

R-PI ~

L, 0 I(t L-, I

CD l-, IQ L..., ..

L..., II >1>

¢- L, ...-...-z·z

',;jW/PP·'2615

0 U'u'0 ~ r-~

I >:T';

f

~ 0, ....

I~I...J

0 :)

1I

+-I

I

I - ,) I

o II I

uj +~l <>

0r, I

0 Q 0 1L_ ~

r J II

_ ,-..J ¢- I

Qr II"

r J I-0

_ ,--1

0 Drill hoI.~ 0 0 0 Dry hoI.-<rt- SAND C i) ProducIng g •• .,."

IlJ • 0 (') 1

0 5000 10.~OOFEET! ,(contour Int.na/, 'Oft, hachur•• Indlcat. clo••d low)

Figure 2·30. Isopach Map, Gatuiia Formation

49

This interpretation would call for movement ofthe Castile halites much earlier than has been suggest­ed. The problem of what caused the movement iscompounded because there would be almost no overly­ing sediments that are needed to instigate the initialhalite movement (the generally classical cause of saltflowage).

The formations overlying the Castile, specificallythe upper two-thirds of the Salado, the Rustler, andthe Dewey Lake Red Beds, show no unusual thicken­ing or thinning across the WIPP site. It is true thatthere are differences in thickness; the Salado thins tothe northwest of the site, the Rustler thins from eastto west, and the Dewey Lake thins east to west. In thefirst two cases, dissolution of halite causes the thin­ning. In the third case, surface erosion is the cause.Nowhere does it appear that these three formationswere deposited on existing highs or lows caused by thevertical movements in the Castile. Even at ERDA-6(Figure 2-4) the above units do not thin, although theyare bowed upward. This indicates that at least some ofthe upward movement of the Castile occurred afterdeposition of the Dewey Lake Red Beds. Additionalevidence presented by Jones (1973) indicates that thisupward movement affects rocks as young as Triassic(Chinle Formation), but not the Ogallala of Plioceneage.

At the WIPP-ll site there is no evidence showingthat the rocks younger than the Castile were pushedupward by the movement of HII. It is possible that the"Infra-Cowden" halite was squeezed out and awayfrom the area of WIPP-ll, but this loss does notaccount for all the extra thickness of the Castile atthat location.

The stretching and consequent thinning of All inthe ERDA-6 area has been discussed by Anderson andPowers (1978). They (and Jones, 1981) have suggestedthat the All unit had been breached by the upwellingHII from below as a diapiric structure. AlII at theWIPP-ll site is thinner than normal and may havebeen thinned by stretching much like All at ERDA-6.It is also possible that a topographic high existed atWIPP-ll during deposition of AlII and that the thin­ness of the unit is caused by original lack of deposi­tion.

On the cross sections (Figures 2-4 and 2-5) All isshown as a continuous band or layer. It is quiteprobable that in the areas of extreme structure, suchas near ERDA-6 and WIPP-ll, the anhydrite consistsoflarger broken and tilted blocks rather than a contin­uous layer.

50

At the WIPP-13 locality AlII is thicker thannormal and HI and HII are thinner than normal. Thearea appears to have subsided prior to or duringdeposition of AlII. If faulting of the underlying BellCanyon Formation and AI of the Castile as shown onFigure 2-5 occurred during this time, some of thehalite could have been forced northeastward underthe WIPP-ll site and AlII would have been depositedin a slowly subsiding local basin.

There is no indication that the Salado and youn­ger beds over the Castile in the WIPP-ll area wereforced upward as they are at ERDA-6. This fact againseems to indicate that the movement in the Castile atWIPP-ll occurred very early in Ochoan time, specifi­cally at or near the end of Castile time. The movementof halite at the ERDA-6 area, in contrast, seems tohave occurred over a longer time period starting dur­ing late Castile and extending to the pre-Pliocene.

The structure and isopach maps of the upper two­thirds of the Salado and younger formations give noreal indication of the large amount of movementfound in the Castile. Structure maps do show theuplifted area around ERDA-6.

Of more interest in the younger formations is thenorthwest-trending synclinal feature in the easternpart of the WIPP site. This feature begins to appearon the structure map drawn on the top of the un­named lower member of the Salado. The structure ismore noticeable on the top of the McNutt potash zoneand the top of the Salado.

Isopach maps of the McNutt and upper unnamedmembers of the Salado do not show a thickenedsection over the syncline. This indicates that subsi­dence occurred after Salado time.

The structure maps of the top of Rustler andDewey Lake Red Beds show that the syncline is stillpresent at its northwestern end, but it dies out south­eastward and actually becomes a closed depression.This apparent disappearance of the plunging synclineis caused by a thick deposit of the Rustler in thesoutheast part of the WIPP site. In this part of thesite, no halite has been removed from the Rustler. TheRustler structure naturally reflects the thicker sectionand the Dewey Lake does the same.

The Santa Rosa Sandstone structure map (Figure2-28) shows no indication of the syncline, but then notmuch of the eastern part of the WIPP site has beencontoured. The isopach map of the Santa Rosa ex­plains the disappearance of the syncline on the SantaRosa structure. There is a thickening of the SantaRosa over the synclinal area. The sedimentary rocks of

the Santa Rosa were deposited in an already subsidedor subsiding trough, and by the end of Santa Rosatime the trough had been filled in.

The subsidence of the syncline began sometimei after Permian time and ceased prior to the Triassic.

The subsidence apparently was rooted in the lowerunnamed member of the Salado Formation.

The Gatufia Formation, a flood-plain deposit, waslaid down about 500,000 yr ago and covered most ofthe area on and around the WIPP site. Remnants ofthe formation are still present, although much of theunit has eroded away. The isopach map (Figure 2-30)shows three prominent deposits on the western half ofthe WIPP site. The two easternmost deposits appearto have been deposited in erosional lows that straddlean east-west high on the Dewey Lake Red Beds. Thishigh is believed to be the result of the Dewey Lake RedBeds having been protected from erosion until recenttime by a cover of Santa Rosa Sandstone.

The caliche and sand-dune cover were studied forevidence of faulting at the surface. Trenches were dugthrough the caliche, but no displacements were seen.Likewise, arroyo banks were examined for displace­ments, but none were found.

From the preceding data it is not possible to datethe age of Castile halite movement, except to say thatit began no earlier than late-Castile time and ceasedprior to pre-Pliocene time.

The synclinal structure apparent in the Saladoand Rustler structure maps began no earlier thanSalado time and ended prior to Dewey Lake Red Bedstime (late Permian). Probably the total developmentof this structure occurred during deposition of theDewey Lake Red Beds.

There is no evidence of movement of any of theCastile or Salado halite at the present time, althoughsome halite at the top of the Salado has been removedby ground water.

2.6 Selected ReferencesJ. E. Adams, "Upper Permian Ochoan Series of Dela­

ware Basin, West Texas and Southeastern New Mexico,"American Assn of Petroleum Geologists Bull 28(11):1596­1625 (1944).

R. Y. Anderson, W. E. Dean, Jr., D. W. Kirkland, and H.I. Snider, "Permian Castile Carved Evaporite Sequence,West Texas and New Mexico," GSA Bull 83:59-86 (1972).

R. Y. Anderson and D. W. Powers, "Salt Anticlines inCastile-Salado Evaporite Sequence, Northern Delaware Ba­sin," in Geology and Mineral Deposits of Delaware Basinand Adjacent Areas, Circular 159 (Santa Fe, NM: NMBureau of Mines and Mineral Resources, 1978), pp 79-84.

G. 0: Bachman, "Cenozoic Deposits of SoutheasternNew Mexico and An Outline of the History of EvaporiteDissolution," USGS J of Res 4(2):135-149 (1976).

G. O. Bachman, Regional Geology and Cenozoic Histo­ry of Pecos Region, Southeastern New Mexico, Open-FileReport 80-1099 (USGS, 1980).

A. L. Brokaw, C. L. Jones, M. E. Cooley, and W. H.Hays, Geology and Hydrology of the Carlsbad Potash Area,Eddy and Lea Counties, New Mexico, Open-File ReportUSGS-4339-1 (USGS, 1972).

C. L. Jones, The Occurrence and Distribution of Potas­sium Minerals in Southeastern New Mexico, NM Geol SocGuidebook, 5th Field Conf (1954), pp 107-112.

C. L. Jones, Salt Deposits of Los Medanos Area, Eddyand Lea Counties, New Mexico (with sections on GroundWater Hydrology by M. E. Cooley and Surficial Geology byG. O. Bachman), Open-File Report USGS-4339-7 (1973).

C. L. Jones, Test Drilling for Potash Resources: WasteIsolation Pilot Plant Site, Eddy County, New Mexico,Open-File Report 78-592, volland 2 (USGS, 1978).

C. L. Jones, Geologic Data for Borehole ERDA-6, EddyCounty, New Mexico, Open-File Report 81-468 (USGS,1981).

C. L. Jones, C. G. Bowles, and K. G. Bell, ExperimentalDrill Hole Logging in Potash Deposits of the CarlsbadDistrict, New Mexico, Open-File Report (1960).

P. B. King, "Permian of West Texas and SoutheasternNew Mexico," American Assn of Petroleum Geologists Bull26(4):535-763 (1947).

S. J. Lambert, "Geochemistry of Delaware BasinGround Waters," in Geology and Mineral Deposits ofOchoan Rocks in Delaware Basin and Adjacent Areas,Circular 159 (Santa Fe, NM: NM Bur of Mines and Miner­als, 1978), pp 33-45.

W. B. Lang, Geology of the Pecos River Between Lagu­na Grande de La Sal and Pierce Canyon, New Mexico StateEngineer 12th and 13th Biennial Rpt (1938).

S. S. Oriel, D. A. Myers, and E. J. Crosby, "West TexasPermian Basin Region," in Paleotectonic Investigations ofthe Permian System in the United States, USGS ProfPaper 515, pp 21-60 (1967).

D. W. Powers, S. J. Lambert, S-E. Shaffer, L. R. Hill,and W. D. Weart, ed, Geological Characterization Report,Waste Isolation Pilot Plant (WIPP) Site, SoutheasternNew Mexico, SAND78-1596, vol I (Albuquerque, NM: San­dia Laboratories, 1978).

R. P. Snyder, "Geologic Data for Borehole WIPP-11," inBasic Data Report for Drillhole WIPP-ll (Waste IsolationPilot Plant- WIPP) , SAND79-0272 (Albuquerque, NM:Sandia National Laboratories, 1982), pp 5-23.

J. D. Vine, "Surface Geology of the Nash Draw Quad­rangle, Eddy County, New Mexico," USGS Bull 1141-B(1963).

51-52

3. Site Deformation

3. 1 Borehole Data

~I mil•

Cowden Anhydrite

______ top Delaware

Mountain Group

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

-===== =====--

_...... -

1000ft •.•.1.-

2000tt •.•.I. -

1000 It b.a.l.-

Figure 3-2. Salado and Castile Formation Log Correlationsin the DZ. (See Figure 3-1 for horizontal distribution ofholes.)

POGO F.d I.,,~ AEC 7

Union F.d I ...... ERDA B •

c---------, 17r"'- - ~ WIPP II • ~, • AEC B

,~ ~,

r J L,rJ L l

I. II IPP

QJI3" WIPP 12 L,

I I

r J I

I ERDA 9 :I II II I

: I~, r J

I IL. _ ~ ,. J

I r J

L "-_.. rJ, r J

L, r--.J..... .J

The DZ was initially identified on seismic reflec­tion lines and confirmed by later drilling. The bore­holes. cores, seismic lines, and regional geologic struc­ture form the data base for the DZ interpretation.

Figure 3-1 shows the location of those boreholes inthe DZ that penetrate the Castile Formation. Figures3-2 and 3-3 are borehole log correlations of the Saladoand Castile Formations along the two lines on Figure3-1. The correlations are tied to sealevel.

Figure 3-1. Boreholes in the DZ That Penetrate the CastileFormation

53

...

Figure 3-3. Salado and Castile Formation Log Correlationsin the DZ. (see Figure 3-1 for horizontal distribution ofholes.)

Within the Salado Formation, the distinctive logcharacter can be used to identify particular strati­graphic horizons. Log correlation is straightforwardand easy. The Salado Formation correlations on Fig­ures 3-2 and 3-3 are convenient wiggles on the com­pensated neutron density logs selected primarily be­cause of the spacing of stratigraphic markers on theselogs. Except for the Cowden Anhydrite in WIPP 11,the units are present on all the logs. The CowdenAnhydrite in WIPP 11 was not cored, and the drill

cuttings over this interval were not described. Basedon the acoustic and density logs, this interval is eithernot present or is greatly altered. Borehole WIPP 11(and to a lesser extent ERDA 6 and WIPP 12) exhibitsa thinning in the lower half of the Salado Formation.This thinning at WIPP 11 is shown in Figure 3-4,which is a tracing of the borehole-compensated acous­tilogs from AEC 8 and WIPP 11 (similar relations areevident on the compensated neutron density logs).The logs are correlated, and the relative thinning ofintervals between markers is plotted along the left sideof the figure.

Within the Castile Formation DZ, the logs can beused to distinguish between halite and anhydrite, butit is generally not possible to identify specific strati­graphic horizons on the basis of log character alone. Inundisturbed areas the logs can be reliably correlatedbecause the units are laterally persistent and changesare gradational. Examples of such correlations aregiven in Snider (1966). Within disturbed areas, logidentification within the Castile Formation may beless certain.

On Figures 3-2 and 3-3, the Castile Formation isdivided into dark and light intervals. The dark inter­vals are primarily anhydrite; the light intervals areprimarily halite. The distinction was made throughthe use of the compensated neutron density logs andmay locally differ from core descriptions or a morethorough multilog analysis. The asterisks indicateoccurrences of brine. All of the pressurized brineencounters in the Castile Formation of the northernDelaware Basin have been in disturbed areas.

Structural deformation is locally intense withinthe Castile Formation (Figures 3-2 and 3-3). There arevertical displacements of hundreds of feet and thick­ness variations of hundreds of percent. In AEC 7 thereare four massive anhydrites with a probable repeat ofAIl; in POGO Fed. 1 there are two (the latter may be adepositional pinchout of HII along the reef margin).In ERDA 6, the uppermost Castile anhydrite unit wasidentified from core as AIl (Anderson, 1976), and AlIIis missing. The sequence of thinner anhydrite bedsabove AlII and in HII is not always observed withinthe DZ.

At WIPP 11 the first massive anhydrite encoun­tered below AU (identified by distinct laminations inthe core) is regarded as AI. If correct, then the top ofAI is -225 ft above its expected elevation. Thisobservation is significant because it implies one of twothings. Either AI is involved in the DZ salt flowagedeformation, or the underlying Delaware MountainGroup is deformed, perhaps by faulting (as interpret­ed in Chap 2).

top Salado Formation

Cowden Anhydrite

top DelawareMountain Group

*

--Sea Lavel-

2000 ft a.B.I.-

1000 ft 8.B.I.-

3000 ft a.B.I.-

1000 ft b.B.!.-

54

40I

AEC8

top CelltileFormetIon

microsecondsper foot

70I

100I

-----------

WIPP 11

I i

Figure 3·4. Acoustilog Correlation of the Salado Formation in AEC 8 and WIPP 11

55

Further indication of some deformation in AI andpossibly the Delaware Mountain Group is described inthe preceding discussion of borehole data from theJames Ranch anticline (Sec 1, T23S, R30E). However,not all disturbed areas are underlain by complexstructure (compare Figures 3-13 and 3-14 in Sec 3.2.2)in the Delaware Mountain Group; the seismic reflec­tion lines show relative continuity of the Cherry Can­yon and Bone Spring events beneath the DZ.

Data from individual wells were used to establishthe geologic section at the borehole. In regions ofsimple, gradational structure, the boreholes and theseismic lines can be used to interpret the actual formof the subsurface structures. In the heart of the DZ thegeologic structure is too complex and the boreholestoo widely spaced to apply these techniques. Eventhough the megascopic structures cannot be interpret­ed, the meso- and microscopic style of deformation isrepresented by existing borehole, seismic, and petro­graphic data. Overall, the DZ is best described as astructural complex.

3.2 Seismic Reflection

3.2.1 DataThe WIPP site was extensively surveyed with the

seismic reflection technique in 1976, 1977, and 1978.These data provide good resolution of the seismic timestructure from the top of the Castile Formation to theBone Spring Formation. Reflections from shallowerand deeper horizons are too unreliable to map withthese data. High-resolution seismic experiments wereconducted in 1979 and 1980 to determine if thesetechniques could resolve the shallower horizons, butresults were generally discouraging. Results of thedetailed gravity survey of 1979-1981 indicate that atleast part of the problem with high-resolution tech­niques stems from lateral variations in velocity anddensity within the formations. These lateral varia­tions are discussed in the following section on theWIPP-site gravity field.

The seismic reflection data were collected duringfour sequential Vibroseis surveys. Table 3-1 lists allthe seismic surveys, including the high-resolution ex­periments. Figure 3-5 is a composite line location mapof the four Vibroseis surveys.

Field and processing parameters and uninterpret­ed seismic sections from the 76SAN, 77X, and 78Ysurveys are in Hern et al (1979). An interpretation of

56

the 78GG survey is given in Bell and Murphy andAssociates, Inc. (1979), but the data are not included.

Table 3-1. WIPP Site Seismic Surveys

Year Designation Line Miles

Vibroseis1976 76SAN 251977 77X 471978 78Y 51978 78GG 74

High-Resolution1979 79EX 11980 80EX 1.3

The following briefly summarizes the field andprocessing parameters used in the WIPP Vibroseisseismic surveys. The 76SAN was conducted by Dress­er Olympic and supervised by Collin McMillan of thePermian Basin Exploration Company (Hern et aI,1979). This survey used 48 receivers at a 220-ft spacingin a 5940-880-0-880-5940 split-spread configuration(processed as a 24-fold CDP stack). The source wasfour vibrators, distributed over 220 ft, generating asweep of 8 to 30 Hz. The signal was sampled atintervals of 4 ms. These parameters are fairly commonto petroleum exploration in the region.

The 76SAN sections do not show reliable eventswithin the Castile Formation, because of the fieldparameters, which were selected to resolve deeperstructures of interest to petroleum geologists. On thesections, the dominant frequency is near 25 Hz andthe vertical resolution should not be better than± 56 ft (10% of one wavelength at 14 000 ft/s). The76SAN sections show fairly consistent parallel eventsover most of the site, the western edge of the DZ nearshotpoint 25 of line 76SAN3, and an area of structuralcomplexity to the southwest of the site (shotpoints 135through 160 of line 76SAN2).

The 76SAN data differ significantly in both reso­lution and seismic character (wiggle shape) from thelater surveys. The 76SAN sections were checked forconsistency with the interpretation of these later sur­veys, but they were not incorporated into the time­structure and isochron contour maps (Figures 3-6through 3-12).

S4 3. 32

-AE£.. 1 l..

51""" T.2IS

r..J_1__ ~ _4_.1---_-

II

7

T. 22 S

6

6

3'

19

18IS

12

1-.,I

14 IL

I'L ! ?AEC 8

~ 'It''

4

~p 11

JJr

6

""'~'$ rII I ." l,-..J 7 a I

'i ~ ! ZONE

10

3

22

27

--

// '\/' ~ II 12 7 • 9 10 \

\

II 12 7 T. 23 S.

II '4

R30E.

l3 16

R31E

II 13 18

R.32 E.

o z •SCALE IN MILES

Figure 3-5. Vibroseis Line Locations. (The three heavy lines are included as Figures 3-13, 3-14, and 3-15 of this report.)

57

-l's

4'0

AEl: 1

310

...t.., I.t.n.h II .1111.....4.4...., .... ft. loLL,..,,,n ..'h" " ...........

"1111I11 TlIIII It'.lt.,.TI, If elltlll Fo'lIIltl,.

....

_+_~--l-_--,,-<,------I_ GG 14

r---,:....-_ 310

.1111.....4. "".... II..

r.1..1. II..• • ··1

" .. If ,1 ,..t ..,..It............,...,----

............--­GIl17--

•)(

II

:s

• Wl...

Figure 3-6. Seismic Time Structure - Top of Castile Formation

,-... IIII.-.--t

58

-I100

-0•• 1111.--1

470

4.0

,,

110

GGI7--

-----..".

'">(

11I1111.....4. t•••••• tllll.111.",1, II••(.....14 •••,. ."'•••••••).... If .",,'11 .t"'t.,.,..It .

..,."1.....,ill'

470

480

___-"'''''- 480

~100 -tS

110

SII.mlo Tim. Struotur.Mid. c•• till Farmltll.

...t ••, 1.,.,..1,10 "'UIl.....4.41"11I' ••1. ft .1.,,.,..14 '" L." ,,. I'"

,." 1

Figure 3-7. Seismic Time Structure - Mid-Castile Formation

59

- ....,.. --1

_71'- .1111_ 'I......... 11.

;;17-- ."•••_ ,. '''.'''111------ ,

Figure 3-8. Seismic Time Structure - Top of Cherry Canyon Formation

60

11'1"'11 ".1 Itr.lt.rI'I, If CllIrr. CII,I. Fer.tI..

ell'''' 1.,.,,,.1'" .1111....._...,••, f' 1.".,.,._ h" " •••. .... _, ,..,

I

1210 -r..1210

1240

1210

'"x

---+-~-t-- GG ,.

l'<

~1250

1210

1240

~O•• MII.~

-1240­

GGI7--

L••••d

"'Ull.....d. Iw•• w•• II",.

,"''''10 Ii ••\......d wh...."'blg••••)

,.••I~I. f••11

Slllmlo Tlml Struot uroTop of Bani Spring Formltion

..., ••, I.,..... , 10 "'11II.....d.d.I.",. 3310 fl.•.•.1.

,ro,...d ~.' Lo"..." •••• Nov.••10ro•. h~.'1

Figure 3-9. Seismic Time Structure - Top of Bone Spring Formation

61

- IG

/,

·...-1-(,(,13i

•IS

~..1·,1

i •

t--......-..-.

;;17-_.

........

LIIIN

"'....... ' 'r.............., ..,. ..., ,. ,-" ,..,.,."1.",1. I....r••f., C..tlI. t. Mil'. CIIUI•

'"','' i.,.,...I: 10 .iIIl••I ••lIf,'r.'.... .., lI"..." •••. In. '"0

Figure 3-10. Seismic Isochron - Top Castile to Mid-Castile

62

tJl

..."

"'7--...........

"'11I1....... I•••••• II",.I~·I.~I.I •• I~I.... '14'1,"''''1' II..1.....14 .NFI ''''~110'''1.... • f ..",,1•• "'0110"

280

!

Slllmlo leaohronMid. Clltll. to Ch.rry C.nyon

alntlur In'trw.l: 10'"1111 ., .., ...4 ~.' L."..." •••• Nov. 1810

Figure 3-11. Seismic Isochron - Mid-Castile to Cherry Canyon

63

1;I

// .10

- -,-,.; ,. 410

oontour IIt',r",1: 10 miltl...."••,r.p.,., '.' larry 1.,'0.1. NI._ "10

Slilmic IlochronChlrry Clnyon to Ionl Spring

- e.c. \7

.,.,.,

XI -----,........<Et-~----~t_-

••

•..

• !ll ... , r-... "t= N

8 "Ii ~

.IO~N><

......--.....11.--1

LoI··'........~ r-IIU.....,. ,••• "'1\1...1'"'·'·1 •• ,.......1•~GI7-- r:."~" u••,...., ..... '''·'1·'·''

Figure 3·12. Seismic Isochron - Cherry Canyon to Bone Spring

64

The nx survey was conducted by Dresser Olym­pic and supervised by John L. Hem of G. J. Long andAssociates. For this and subsequent surveys, fieldtechniques were modified from those of the 76SANsurvey to obtain better resolution of the shallowerhorizons. The nx, 78Y, and 78GG surveys used24 receivers at a spacing of 110 ft in a split-spreadconfiguration (processed as a 12-fold CDP stack).Vibrator sweep frequencies were increased to 25 to100 Hz, and the sample rate was increased to 2 ms.The nx and 78GG surveys both used a 1650-440-0­440-1650 split-spread configuration with three to fourvibrators (depending on how many were operational)stepped over 220 ft. The 78Y survey used a 1440-110­0-110-1440 configuration with one vibrator steppedover 110 ft.

The nx sections indicate the seismic features ofinterest at the WIPP site and extend the data beyondWIPP site Zone III. By themselves the nx lines weretoo widely spaced to allow a reliable three­dimensional interpretation. They were worked intothe interpretation of the 78GG sections and form partof the data from which the following maps were made.

The 78Y survey was run by CXC Inc. and pro­cessed by Seismic and Digital Concepts, Inc., both ofHouston, Texas. John L. Hern of G. J. Long andAssociates managed the survey.

The field techniques were similar to the nx and78GG surveys, except that the 78Y survey used onlyone vibrator instead of three to four and a shorternear-offset distance (110 ft vs 440 ft). The processedsections show reliable events, but they are not as clearas those from the nx and 78GG surveys. The rela­tively better quality of data from these other surveysis likely due to the repression of surface noise bymultiple vibrators.

The 78Y lines coincide with some of the 78GGlines. The 78GG sections were used in the seismicinterpretation because of the better apparent qualityof data and because the redundant 78GG sections arepart of a larger data set. The interpretation waschecked against the 78Y sections; no inconsistencieswere found.

The 78GG survey was conducted by Grant Geo­physical Corp. (formerly Dresser Olympic) for BechtelNational Inc. and supervised by Bell and Murphy andAssociates, Inc. (1979). Twenty lines cover WIPP siteZones I, II, and III at an approximate spacing of1/4 mi. Field parameters were identical to the nxsurvey, except that the vibrators used an upsweep(frequency increasing with time) instead of a down-

sweep. The 78GG survey also included two experimen­tal lines (78GG5X and 78GG6X) with a receiver spac­ing of 55 ft and a 935-330-0-330-935 split-spreadconfiguration. The processed sections of these experi­mental lines revealed no data that were not alreadycontained in the normal sections.

3.2.2 InterpretationThe 78GG survey forms the basis for the present

interpretation. The nx sections were used to verifyand expand this interpretation.

The processed seismic sections, along with theboreholes, core descriptions, and regional geology,form the basic data set for interpreting subsurfacestructures at the WIPP site. It was thought that thegravity field would provide significant additional con­trol on the interpretation. However, the gravity fieldwas found to be dominated by lateral density varia­tions in relatively undeformed shallower strata; gravi­ty data are interpreted in a later section.

Figures 3-13, 3-14, and 3-15 are three representa­tive interpreted seismic sections. Locations of linesnX5, 78GG3, and 78GG19 are indicated on Figure3-5. Four interpreted seismic horizons and the tie toboreholes WIPP 12 and WIPP 13 are indicated on thesections. (The four interpreted horizons are identifiedlater in Table 3-3.)

The interpretive process involves more than iso­lated seismic sections. Each seismic horizon is initiallyidentified with one of the continuous series of troughson the individual traces (vertical wiggly lines) thatmake up a total section. The horizons are carried alongthe section to intersecting cross lines. The two-waytravel times are transferred to the intersecting section,and the horizons are carried along this section to thenext intersecting cross line. This process is repeateduntil the horizons are tied back to themselves on theoriginal section. In this way the interpretation is de­veloped as a system of closed loops. Seismic two-waytimes are read off the sections, posted on shotpointmaps, and contoured. Structural features (e.g., faults)are correlated between adjacent or intersecting sec­tions, located on the shotpoint maps, and worked intothe contours. The incremental two-way time betweenhorizons is also posted and contoured as isochrons.Finally, the resulting time-structure and isochronmaps are examined for geologic consistency. The mu­tual cross-checks of the entire interpretive processresult in maps that are more reliable than could beachieved with the sum of the individual pieces.

65

66

Figure 3-13. Seismic Line 77X5

..LBS HEORNli!IS

LINE X~5

SS~~TT~~::S"'}:E3;~i~

~'~'iHi. ..

~·~T""""

..~._-----------------

PR0CESS I NG SEOUENCE

II~I •....~f ."-11:'_.1_, .

';'I_.I.~

':.I_I'~I••:~_"lll'l-'"

GRANT.m

" __'~,""IQ

" .....,"-.....

-,\-.._,_,t .. ~1_", _ ._

u.s. DEPT. BF ENERGYWIPP LINE 3

CBNVENTIBNAL STACKSTRTI0NS 101-269

BECHTEL J0B N0. 12484-002

~ELEv,:lT 18HS

STRTleNS

V£Lec IIf FUNCT leND(RECT 18"1

1'tL.QIRfCTl8N

'-'1::::1::.- -"....

3

'.'~

.05

~~

.~..~

~ ~~.':J 311I

110

....

I~

.'

'25

t

'I"

'fJ1~J':rI V "1'1"'-I I

• 0 . • ~

_ ..

'L_ _

• •I",,.',0

? ....•'>I..

\70'75'10.'"

"" "'..~

'10

"~

~

,..Imlmuumr .. .MlnJ .U!Jllll

' .... ' .-

...

"~, •."-'"'. '.~ ~ - .

,:;~ '... '.~ I' . " -- ' .

(,_.~

, • • ,,t•••• ,...

,..~TT ~~~ !

,rl,,, = 7" 7'0 '30 'I" ,. ,~ 2. 20 2l" ~I"

til•r

4·"~~.'.·'~~tL~ L't ((;', .. ~:>'. ',';.

67

'tW''.;, .

''- e..-'J--. ,",_

..]~

Figure 3-14. Seismic Line 78GG3

....... ,~

I

... ,. -1-

.:t

"'11I

.. ,:1,

"<c~••• '

", "

."..,.:_ .. ,c:.

-..-:_._ ....JM:,._

"1.'''-''':''0'''~ ...:,-....,-,.,...:_.-

_0'_I ., .......-- --,- - - ,-

GRANTSEW

PReCESSING SEQUENCE

u.s. DEPT. 0F ENERGY~JPP LINE 19

C0NVENTI0NAL STACKSTAT IBNS 101-309

BECHTEL jaB "lB. 12_84-002

.......,_ ...." --..........,tt_.,__':"_J'~'",,-_ ..,....­,:_,._ ......: ..0 ..

_'..'11.01..-.: .....

" :'::.~:-:-.::'::: -.otT ~ .. _

,,_..........,..

STATI0N5

vELee I T'J FUNCT 10NDI~ECT leN

! [hlLOIRE'! 10N

~:ELEVAT[0NS

2

3

4

------------~----------------------~-----------__cc

----~------~~---- -------~~~~~~~~~~~~~~~~~~~~~~~~~~~~-==

., -

-i·1

~I."

Figure 3-15. Seismic Line 78GG19

68

The mutual cross-checks also allow the interpret­er to develop an understanding of the quality of data.The WIPP seismic time structure (Figures 3-6, 3-7,3-8, and 3-9) and isochron maps (Figures 3-10, 3-11,and 3-12) are regarded by this interpreter (LarryBarrows) as reliable.

Several features on the seismic time structure andisochron maps warrant further discussion. First, mostof the site is characterized by broad, open, low­amplitude (10 to 40 ms of closure) folds. Except whereshown on the maps, the events are continuous, indi­cating the absence of significant faults. Over most ofthe site the isochron maps are featureless, indicatinguniform stratigraphic thicknesses.

There is a northeast-trending fault southeast ofWIPP site Zone III. This feature is strong on linesnX3, nX6, 78GG11, and 78GG19; moderate on line78GG10; and weak to nonexistent on lines 76SAN1,nX5, and 78GG18. It appears to extend through thetop of the Castile Formation and down into the under­lying Delaware Mountain Group. This displacement issmall (10 to 20 ms) and shows no consistent offset.The arcuate map trace, lack of consistent offset, andabsence of a fault indication at its projection onto linenX5, all suggest this is not a major through-goingtectonic feature. Figure 3-15 shows the feature on line78GG19.

There is an east-trending elongated syncline inSec 29, 30 of R31E, T22S. It is best seen on the twoCastile Formation time-structure maps (Figures 3-6and 3-7). The relief is -30 ms two-way time (210 ft at14 000 ft/s). The sides of the syncline appear to beslightly (5 to 10 ms) faulted. This fault indication canbe seen on line 78GG3 (Figure 3-14). If the faultindications are valid, this feature is a small graben.

The largest anomaly on the seismic sections is theDZ. Seismic line nX5 (Figure 3-13) is reasonablyrepresentative of data from this area. This line ex­tends from a normal area across the DZ and back intoa normal area. From shotpoints 112 through 160 theCastile horizon reflections are considered too unreli­able to map. The seismic data are valid, but thegeologic structures within the Castile Formation aretoo complex to map with the seismic technique. Theseconclusions are supported by the steep laminationdips, variable stratigraphic thicknesses, and petro­graphic features (e.g., recumbent folds, shear zones)exhibited by core. The seismic data indicate a blockystructure with abrupt dip changes and offsets (faults)between units. The seismic character (wiggle shape)

changes, which indicates variations in thicknessesand/or acoustic properties. From shotpoints 160through 190, Figure 3-13 indicates an anticlinal flex­ure on the mid-Castile horizon.

The use of seismic bright-spot techniques to de­tect brine pockets within the Castile Formation hasbeen suggested. These techniques are based on thechange in acoustic properties accompanying changesin the content (gas, oil, water) of pore spaces. Suchacoustic variations sometimes cause changes in seis­mic character that can be correlated with a prospec­tive geologic structure that indicates an oil or gasreservoir. Based on core examination, the WIPP 12brine occurrence is from a few nearly vertical fractureswithin AlII. These fractures are not expected to pro­duce an identifiable seismic signal. If the fractures arepart of a spatially finite, pervasively fractured volumeof anhydrite, the associated variations in bulk acousticproperties might produce a change in seismic charac­ter. This change would have to be distinguished fromthose resulting from the existing but unknown lateralvariations in stratigraphic thicknesses. At this timethe use of bright-spot techniques to detect brine pock­ets does not appear feasible.

The lateral extent of the DZ is best establishedfrom Figures 3-6, 3-7, 3-10, and 3-11. These mapsshow an "area of complex structure" within which theCastile Formation's geologic structures are too com­plex to map with the seismic technique. This areaincludes boreholes WIPP 13, WIPP 11, ERDA 6, AEC7, and is open to the northeast. Bordering this is anarea where the Castile Formation horizons show saltflow structures and some faulting. Boreholes AEC 8,WIPP 12, and SCL Fed. No.1 are in this peripheralregion. The outer edge of the mapped flow structuresis here taken as the limit of the DZ. This definition isnecessarily ambiguous. It includes the anticline atWIPP 12, and mayor may not include the anticline inSec 19.

The seismically indicated DZ affects primarily theCastile Formation horizons. The underlying CherryCanyon and Bone Spring horizons have reduced dataquality, but are continuous enough to map. As notedbelow, the irregular seismic time structure on thesehorizons may be caused by velocity variations withinthe evaporites. Borehole control supports the inter­pretation that the DZ structures either do not extendinto the Delaware Mountain Group or are very muchreduced in amplitude.

69

Table 3-2. Stratigraphic Identification ofUppermost Seismic Horizon

The deeper seismic horizons are only tentativelycorrelated with the stratigraphic section as indicatedin Table 3-3.

section 78GG20 indicates that the top of the CastileFormation is at or very near the uppermost seismichorizon.

3.2.3 Stratigraphic CorrelationThe seismic horizons need to be identified with

particular levels in the stratigraphic section. At bore­holes the stratigraphic section is known in feet; seis­mic horizons are known in seconds of acoustic traveltime. Uphole velocity surveys are used to correlate thetwo.

Uphole velocity surveys are available for ERDA 9,WIPP 12 (above 2789 ft), WIPP 13, WIPP 18, andWIPP 34. Additional surveys in WIPP 11, WIPP 12(below 2789 ft), WIPP 14, and WIPP 33 would aid thegravity interpretation. The new WIPP 12 surveywould also help identify the mid-Castile seismic hori­zon.

The uppermost mapped seismic horizon is nearthe top of the Castile Formation. Boreholes WIPP 18and WIPP 34 do not penetrate to this depth. WIPP 13is within that area of the DZ where geologic structuresare too complex to map with the seismic technique.The geologic markers at this borehole are indicated onthe seismic sections (Figures 3-13 and 3-14).

The uphole velocity surveys of WIPP 12 andERDA 9 were used to determine the acoustic two-waytravel time to the top of the AlII member of theCastile Formation. Depths and travel times are indi­cated on Table 3-2. Posting these times onto seismic

Kelly BushingGround LevelDepth to Anhydrite III(from KB)Time to 3300 datum(from survey)Time to KB(from survey)Adjusted two-way time to3350 datum (@ 6000 ft/s)

ERDA 9(ft)

342034092836

0.2000

0.417

WIPP 12(ft)

348434722741

0.2114

0.378

Table 3-3. Tentative Stratigraphic Identification of Lower Three SeismicHorizons

Depth at Seismic TimeSite Center at Site Center Interval

Seismic Stratigraphic (GCR Fig. 3.3-2) (3350-ft datum) VelocityHorizon Identity (ft) (s) (ft/s)

1 TIAnhydrite III 2825 0.41714200

2 Anhydrite II 3450 0.50512200

3 TICherry Canyon 5100 0.77512100

4 T/Bone Spring 8000 1.255

70

Figure 3-16. Comparison of Interval Velocities MeasuredDuring Uphole Velocity Surveys - ERDA 9, WIPP 12,WIPP 13

WIPP 12. The conclusion is that the lateral velocityvariations preclude reliable conversion of the seismictime structure and isochron maps into geologic depthstructure and isopach maps. The seismic maps reli­ably display seismic time structures, but it is uncer­tain whether the indicated time structures result fromgeologic depth variations or seismic velocity varia­tions.

In petroleum exploration, seismic maps are al­most always converted to geologic depth maps todetermine the extent and closure of prospective struc­tures. If no other information is available, a standardvelocity is used in these cases to make the conversion.For the WIPP site evaluation, variations in lateralvelocity may be as significant as the geologic depthstructure. Because there is not enough information toseparate the two effects and because both may besignificant, the seismic interpretation is left in theform of seismic time structure and isochron maps.

i iL. I

: ~- --------------,L._._._._._._._._._. .• I

WIPP 13..../ : ERDA 9""":. II Ii L,

I I

i :i Ii I

r JL._._, I

i Ii II I

! :"'--l

WIPP12/ :I

17

SaladoFormation

1615

Interval Velocity

Ix1000 ft/a)

14131211

o

3000

1000

3.2.4 Velocity ControlThe seismic interpretation resulted in a set of

time-structure and isochron maps. Converting theseseismic maps into geologic depth-structure and iso­pach maps requires a known (or assumed) velocitystructure.

The velocity structure at the WIPP site is knowndirectly through uphole velocity surveys and is indi­rectly indicated by the gravity field (see Sec 3.3). Bothdata indicate large lateral velocity variations withinthe stratigraphic layers.

The uphole velocity surveys measured the acous­tic travel time between a source at the surface andreceivers at known depths in the boreholes. TheERDA 9, WIPP 12, and WIPP 18 surveys were con­ducted by Seismic Reference Service, Inc. by usingexplosives in shallow (about 100-ft) shotholes. TheWIPP 13 and WIPP 34 surveys were conducted by theSeismograph Service Corp. through the use of me­chanical vibrators on the surface of the ground.

It is difficult to compare the source-to-receivertravel times between these different surveys becausethe field parameters (seismic wavelets, near-surfacevelocity variations, depths of measurements) differ.However, the interval velocities are found by dividingthe depth increments by the corresponding incre­ments of travel time; these interval velocities can thenbe compared directly.

In the following section on the interpretation ofgravity, the interval velocities of the Dewey Lake,Rustler, and upper part of the Salado Formationmeasured at WIPP 13 are contrasted with those ofWIPP 34. It is shown that the seismic velocity in theRustler and Dewey Lake Formations at WIPP 13 issignificantly less than at WIPP 34.

These variations of velocity apparently extendthroughout the Salado Formation. Figure 3-16 is acomparison of the interval velocities measured atERDA 9, WIPP 12, and WIPP 13. The surveys dif­fered in the number of measurements; ERDA 9 hadthe fewest. For this comparison, the interval velocitiesfor WIPP 12 and WIPP 13 were recalculated at incre­ments similar to those for ERDA 9.

The contrast in velocities for the Salado Forma­tion between ERDA 9 and WIPP 12 accounts for20 ms of seismic time structure. Similar effects resultfrom the contrast for the Rustler and Dewey LakeFormations between WIPP 13 and ERDA 9 and

71

The broad open folds may be caused either bygeologic depth structures or by lateral velocity varia­tions. Their interpretation is ambiguous. However,the subtle fault indications (5 to 20 ms) discussed inSec 3.2.2 cannot be attributed to lateral velocity varia­tions. This follows from the field and data processingparameters used to prepare the record sections. Recallthat the survey used a 1650-440-0-440-1650 split­spread configuration with receivers spaced at 110-ftintervals. This means that receivers were located atdistances of 440 to 1650 ft on either side of theshotpoint (24 receivers). For each sequential shot theentire assemblage was advanced 110 ft along the line.(Physically one new receiver is added to one end, oneis removed from the other end, the vibrators advance110 ft, and receiver outputs are assigned new identifi­cation numbers.) During 12-fold CDP data process­ing, the receiver outputs are gathered into groups ofshot/receiver pairs where each pair in the group issymmetrically distributed with respect to the com­mon-depth-point (CDP) in the middle. Each gather isadded or stacked into a single trace in a method thatemphasizes the best velocity for the reflections. De­tails of CDP shooting are described by Waters (1978).A single trace on the processed seismic section iscomprised of energy traveling along 12 different pathsdistributed over 1650 ft of the line. An abrupt changein the velocity of the overlying rock will gradationallyaffect 30 (1650/110 x 2) traces. However, small faultsaffect only a few adjacent traces. The resolving powerof seismic reflection data is much better when consid­ering small offsets between adjacent traces than whenmeasuring total depth to geologic strata.

3.3 GravityThe subsurface geologic structures at the WIPP

site are known primarily through boreholes and seis­mic sections. These data indicate broad open foldsover most of the site; a northeast-trending fault in thesoutheast corner of Zone IV; an east-trending faultedsyncline (possibly a graben) in Sec 29, 30 of R31E,T22S; an area of complex structure in Sec 1 2 of R30ET23S, and an area of complex structure in' the north~ern part of the site (the DZ). Most of the deformationis restricted to the Castile Formation and involvesredistribution of the massive anhydrite and halitemembers of this formation.

IThe density stratification of the Castile Forma-

tion is particularly striking. Figure 3-17 is a copy ofthe compensated neutron density log of the CastileFormation in borehole AEC 8. The three massiveanhydrite members with a density -2.95 g/c3 areseparated by halite members with a density -2.1 g/c3

•

The strong density contrast between the massiveanhydrite and halite members and the localization ofdeformations to the Castile Formation suggested useof the gravity technique to investigate the structures.A survey was planned, and the gravity effect of somepostulated structures was modeled to assist in select­ing the field parameters. The modeled gravity anoma­lies were generally small (a few tenths of a milligal)and had double half-widths -2 km. The WIPP surveywas originally planned to resolve these anomalies.

_ IIAlADO _

FORMATIONANHYDRITE III ANHYDRITE II ANHYDRITE I BEUCANYON

FORMATION

j

~

o..:ac~

:~C'lil"!l!l..ooj

Figure 3-17. Compensated Densilog, Castile Formation, Borehole AEC 8

72

SNLA had purchased a petroleum explorationregional gravity survey of the northern Delaware Ba­sin before conducting the WIPP survey. (These dataare proprietary and cannot be released.) This surveyprovides control for the regional gradient at the WIPPsite and indications of an anomalous gravity field inthe area of the DZ.

The WIPP gravity survey covered all of Sec 8, 9,16, 17, 20, 21, 28, 29 and the southern third of Sec 4and 5 of R31E, T22S. Stations or observations werespaced at 1/18-mi intervals along north-south lines1/6 mi apart. Figure 3-18 shows the area of the surveyand a net of reconnaissance profiles in the vicinity ofborehole WIPP 33.

The gravity field over the WIPP site is dominatedby a strong regional gradient associated with theDelaware Basin and by a broad, east-trending, posi­tive 1-mgal anomaly that peaks near the line betweenSec 20, 21, and 17, 16. This positive anomaly wasindicated by the proprietary regional gravity survey.It coincides with a topographic ridge (Figure 3-18) andwith a seismic time-structure anticline in the Dela­ware Mountain Group (Figure 3-8). The correlation issuch that 0.05 s of seismic two-way time structurecorresponds to a 1-mgal gravity anomaly. Both thebroad seismic time structure and the associated gravi­ty anomaly could be due to geologic depth structuresand a normal density increase with stratigraphicdepth of about 0.22 g/c3

• However, borehole depthsindicate deep structure that is discordant with thisinterpretation. An alternate interpretation is that thebroad seismic time structure is a velocity effect result­ing from lateral velocity variations within the overly­ing strata. The higher velocity material overlying theseismic pullup should have higher density that, inturn, would cause the positive gravity anomaly. Later­al velocity variations sufficient to cause such seismiceffects are indicated by comparison of uphole velocitysurveys through the Dchoan Series.

The WIPP gravity data were reduced by subtract­ing both a linear regional trend and a simple second­order polynomial surface approximating the broad1-mgal anomaly. Neither of these regional fields isrigorously controlled by geologic data; thus the residu­al map should be regarded as having undeterminedfirst- and second-order trends. The elevation correc­tion density of 2.3 g/c3 was selected from boreholedensilogs and inspection of gravity profiles over thetopographic hill in the southeast corner of Sec 28.

Figure 3-19 is the resulting simple Bouguer gravitycontour map.

The features on Figure 3-19 differ significantlyfrom those anticipated during planning of the survey.The anomalies are much too sharp (shorter doublehalf-width) to originate within the Castile Formation.They extend into areas that are indicated by theseismic profiles as undeformed.

The negative gravity anomalies were establishedby drilling to originate from lateral density variationswithin relatively flat strata. A detailed survey wasconducted over a topographic depression that may bean alluvial doline centered on one of the negativegravity anomalies. The area of this detailed survey isindicated on Figure 3-19 and the resulting simpleBouguer gravity map is shown on Figure 3-20. Thedata for this detailed map were reduced with the sameelevation correction density and regional trend asused for the WIPP survey map (Figure 3-18).

Figure 3-21 is a profile across the anomaly alongthe line indicated on Figure 3-20. The anomaly at thislocation is -0.4 mgal and has a double half-width of500 ft. If a two-dimensional structure is assumed, thetop to the causative density structure should be at orabove 250 ft (Nettleton, 1976, p 192), and the mini­mum missing mass was calculated as 470 tons perlinear foot along the anomaly (method described inNettleton, 1976, p 212).

Also shown on Figure 3-21 are the family of hori­zontal cylinders of varying densities and radii thateach causes the modeled gravity anomaly. They showonly the approximate scale of possible density struc­tures and are not intended to indicate an interpreta­tion.

Borehole WIPP 14 was drilled at the locationindicated on Figure 3-20. The hole encountered nor­mal depths to the stratigraphic horizons, but themeasured densities in both the Dewey Lake and Rus­tler Formations were less than at nearby WIPP 34.Part of the decreased density could be caused byconversion of anhydrite to gypsum.

The negative gravity anomaly at WIPP 14 extendsgenerally west-northwest across Sec 9 and 8. Seismicline X2 runs north-south through the site center andtransects the gravity anomaly near shotpoint 90. Theshallow seismic events near the Rustler Formationexhibit a pronounced depression in the area of thegravity anomaly. Both the seismic time structure andthe gravity anomaly can result from lateral variationsin density and velocity.

73

" " ' _TA.TF~\:,;:'1:.:.. , uF rl-iE ,NTEl{lCHr

GEULOGICAL SURVEYc....~-::.:C'~... ,:;..•• ""' ... ,.,.

TopacraphrbJ"-J,GlIihnIr,R.O,DM,N,C.~. C. E. W...., R. E.IAo. R. E, "'-'.N, 8. WriIrl, J. A. .....,H~~,Mld~PtIcosR_s..n.,.

~~~!~-;:t~" _,.,_,......._l.~ T~,.,__... _ ,__

SCN.El62!lC1)

~~J~~_~'_-_-~'=_.-..._ 0 _ _

.-:..-~-~',..,

Contour latern.l 10 leftDMoo••__ .."

FCNI SAU .... U S gEOI.DGlCM SUIN'£Y. DUIV(II. COl.-CNlI\OD 1Nl' Oft "ASHINCl1"DfII, 0 c. aazu.~OLOl.Ol_IIIlI~_ D _ llOfO"OUl.'

N[W ME~ICO

,EDDY COUNT"

NA.SH DRAW QUADRANGLEIS-MINUTE SERIES

........._, 1"'_-'-'_IOOOO_ ........... ..ol .....,-----:==l~=,:·:::-.. _·... .,..

NASH DRAW, N. MEX1Il1215-..I0J4~'I'

,..._1"__01.'.... _,,,-......,..,

Figure 3-18. WIPP Gravity Survey, Detailed Areas, and Reconnaissance Profiles

74

FIG. 2 1-4. MAIN SITE SURVEY

Simple BougueLe.. Llna., ,GnvltvSlab Regional • PDenilly 2 3 araboll<: T ~0.05 MIIII•• , c'on:m/CC undContours Within t our Interval

F" 0.025 m.IQure 3-19. WIP . 0 I ,P Gravity Surv S' -1000 It.ey, Imple Bouguer Gravity Map

75

•

L4--~--

a

.' , ~---

~--~-~-/

-----,~-\

'----..

•

.~ Detail Gravity Survey - WIPP 14Simple Bouguer Gravity

slab density 2.3 g,%cIinear regionalc.i.D,D2 milligals

Doline indicated by dashed contours• Gravity Stations

Figure 3-20. WIPP Gravity Survey, Simple Bouguer Gravity Map in Vicinity of Borehole WIPP 14

76

AlOu1hwest

•~1 modeled0.1 mgeli

surface

1100'~

@),2.3 glee

Figure 3·21. Profile Across the Negative Gravity Anomalyat Borehole WIPP 14

Figure 3-22 is a copy of seismic line X2 betweenshotpoints 58 and 105 along with seismic two-waytimes for borehole WIPP 11 (Table 3-4). WIPP 11lacks the uphole velocity survey needed to rigorouslyconvert horizon depths to seismic times. An approxi­mate tie was made with measured velocities fromWIPP 34 for rocks above the Salado Formation andfrom WIPP 13 for rocks below the Rustler Formation.This combination was used because WIPP 13 lies in ashallow-origin negative gravity anomaly; both WIPP11 and WIPP 34 are in a normal gravity field. Veloci­ties for the deeper horizons are from WIPP 13 becausethis is the closest borehole with measured velocities inboth the Salado and Castile Formations.

Gravity survey line G (Figure 3-23) runs north­south through the site center coincident with seismicline 77X2. The anomaly is identified in Figure 3-23,removed from the indicated broader trend, and plot­ted in Figure 3-22 on the same horizontal scale as theseismic section.

The seismically indicated syncline extends be­tween shotpoints 80 and 97, has a maximum ampli­tude of 0.035 s, and is generally at the level of theRustler Formation. The gravity anomaly at this loca­tion has an amplitude of -0.55 mgal and a doublehalf-width of 1400 ft. The top of the density structurecausing this anomaly should be no deeper than 700 ft.

If the seismically indicated syncline is a structuralfeature with no associated lateral velocity variations,it would have -135 ft of closure (0.035 s two-way timeat 7734 ft/s).

If the seismically indicated syncline is caused bylateral velocity variations in rocks above the RustlerFormation, the necessary velocity contrast is 7734 to6420 ft/s.

662 ft at 6420 ftls = 0.206 s two-way time662 ft at 7734 ftls = 0.071 s two-way time

~ = 0.035

The high velocity (7734 ft/s) was measured at WIPP34. The low velocity (6420 ft/s) compares reasonablywith the 6549 ftls measured at WIPP 13.

Table 3-4 shows the interval velocities measuredat WIPP 13 and WIPP 34, along with the seismic two­way times calculated for WIPP 11. Figure 3-24 com­pares the interval velocities calculated from the WIPP13 and WIPP 34 surveys. The lateral velocity varia­tions are consistent with the observation that WIPP13 lies in a negative gravity anomaly.

The empirical Nafe and Drake curve (Nettleton,1976, p 252) relates rock densities to velocities. TheNafe and Drake relation works well for common sedi­mentary rocks undergoing normal compaction andlithification. It is less reliable for evaporite rocks andshould be applied with caution. For a velocity contrastof 7734 to 6420 ftls, the corresponding density con­trast is near -0.12 g/c3

• Figure 3-25 is a model withthis density contrast and a maximum thickness of200 m (656 ft). The details needed to fit the model tothe gravity observations are within the uncertainty inthe control.

A first-order approximation of the seismic effectcaused by the density structure is made with thefollowing assumptions:

• The density contrast of -0.12 g/c3 is associatedwith a velocity contrast of 6420 to 7734 ft/s.

• Vertical ray paths to the top of the RustlerFormation from coincident source/receivers atthe surface.

This approximate seismic effect is shown in Figure3-22.

A better seismic model would include the effectsof ray bending at the sloping interface of the anoma­lous structure. Another, probably more important,correction is the effects of the actual vibratorlreceiverarray. Note that the total width of the modeled body is550 m (1804 ft), and the survey used a 1650-440-0-440­1650 split-spread configuration (processed as a12-fold CDP stack). Note also that the signal wasmuted so that shallow events are from only the innerreceivers; deeper events are from all receivers. Thisprobably explains why the shallow seismic structure isconsiderably reduced at the deeper horizons.

77

t0.2 mi II igal

+

• • • •\

•

\•

\ .."._.J

•

/•

WIPP IIgJ. 3426

• • • • •• •

Rustler

I a 9 9 8 7 7 65 6

Figure 3-22. That Portion of Seismic Line 77X2 That Intersects the Negative Gravity Anomaly

78

Table 3-4. Meas\jred Velocities and Calculated WIPP 11 Depth-Time Tie

WIPP 34 WIPP 13 WIPP 11(GL 3433 ft) (GL 3405 ft) (GL 3426 ft)

assumed Seismic 2-Way TimeHorizon depth time velocity depth time velocity depth velocity time (datum corrected to

Top (ft) (s) (ft/s) (ft) (s) (ft/s) (ft) (ft/s) (s) 3200 asl @ 6000 ft/s)

Surface -0.07537734 6549 7734

Rustler Formation 652 0.0843 518 0.0791 662 0.0856 0.095912570 11 474 12570

Salado Formation 965 0.1092 845 0.1076 943 0.1080 0.140714426 14426

Anhydrite III 2957 0.2540 2330 0.2041 0.332919266 19266

Halite II 3508 0.2826 2411 0.2083 0.341315600 15600

Anhydrite II 3625 0.2901 3377 0.2702 0.465118958 18958

Halite I 3716 0.2949 3499 0.2767 0.478114921 14921

Anhydrite I 3810 0.3012 3549 0.2800 0.484716428 16428

TD 3856 0.3040 3567 0.2811 0.4869

Figure 3-23. Gravity Survey Line G

1.1.141210a•

WIPP13~

42

f2001<

!

- TWIPP34 I

..... L-- - - - ----- ----1 ,-WIPP34WIPP13 T L ,

IIII

r"II

r -----'III

" r·........ \ ... ,...~( :. \... \ \,.... Iocal regional

1 \~./ ~.: • anomaly~ i..

~i ..• \ /..•• .\t.

V

I soc 28/29 I 88C 20/21 I _ 18/17 I soc 8/9 I 88C 4/5

South1.0

o

0.8

-0.2

f 0.8

~= 0.4

Ift 0.2

~

._Voloctty

Cx1000ft/ol

Figure 3-24. Comparison of Interval Velocities MeasuredDuring Uphole Velocity Surveys - WIPP 13, WIPP 34

South

• •

l0.1 mgal

.i

•

•

North

•

-.12 g/cc

-surface-

Figure 3-25. Modeled Gravity Anomaly - Survey Line G

80

The seismic and gravity models show that theobserved seismic and the negative gravity anomaliesmight originate from common lateral variations invelocity and density. The possibility of structuralsynclines or shallow stratigraphic channels cannot beruled out, but appear very unlikely considering theexisting borehole control, well logs, and core descrip­tion.

The existence of the lateral variations in velocityand density is reasonably established by the gravity,seismic, and borehole data along with the upholevelocity surveys. A common consensus about theirorigin has not formed among the WIPP geologic inves­tigators. L. Barrows believes they result from karstprocesses. The following discussion is of that interpre­tation by L. Barrows and is not necessarily agreed toby others and does nut imply endorsement by SandiaNational Laboratories.

The detailed gravity survey midway along the linebetween Sec 9 and 16 (later drilled by borehole WIPP14) was centered on a closed topographic depression.This depression is -10 ft deep and 700 ft across, andit is one of many scattered through the Permian Basin.It is interpreted as an alluvial doline formed whenloose surficial material washes through cracks in theDewey Lake Formation into solution conduits in theRustler Formation.

Borehole WIPP 33 was also drilled in a closedtopographic depression and encountered normaldepths to the stratigraphic horizons. This depressionand borehole are in Sec 13 of T22S, R30E -1 mi eastof the edge of Nash Draw. A network of reconnais­sance gravity profiles shows a negative anomaly of0.6 mgal, with a double half-width of 900 ft centeredon this depression.

Borehole WIPP 33 encountered four cavities inthe Forty-niner and Magenta Dolomite Members ofthe Rustler Formation. The alluvial dolines are rea­sonably related to cavernous zones, and the correla­tion of negative gravity anomalies with the dolinessuggests that the lowered velocities and densities arealso related to cavernous zones. The anomalies are toolarge to be due solely to the open space of the cavities.However, karst channels are persistent over long timeperiods, and it is reasonable to think that rock nearkarst channels would have altered petrophysical prop­erties, perhaps by leaching, and that the anhydritewould be hydrated to gypsum. The negative gravityanomalies would then result from decreased rock den­sities near karst channels, primarily in the Rustler

Formation. This interpretation is consistent withthose given by Arzi (1977) and by Omnes (1977) formicrogravity surveys in other karst regions.

The interpretation of the negative gravity anoma­lies is understandably ambiguous. Spatial correlationbetween the negative gravity anomalies, the closedtopographic depressions (here interpreted as alluvialdolines), and the cavernous zone in WIPP 33 may becoincidental. Not all of the topographic depressionshave negative gravity anomalies, and not all negativegravity anomalies have closed topographic depres­sions. The seismic time-structure syncline on line77X2 would be considered unreliable if it did notcorrelate with the negative gravity anomaly. Addition­al uphole velocity surveys (WIPP 11, WIPP 14, andWIPP 33) and borehole gravimeter surveys (WIPP 12,WIPP 13, WIPP 14, WIPP 33, and WIPP 34) arerecommended.

Particularly important is the petrologic nature ofthe lateral variations within the Dewey Lake Forma­tion. Inspection of the core from WIPP 14 and com­parison with core of the Dewey Lake Formation inWIPP 19 did not reveal obvious differences. Furtherpetrophysical analysis might provide additional in­sight into these lateral density variations.

Stratigraphic facies refers to aspects of the lithol­ogy that are attributable to lateral variations in thedepositional environment. The possibility that thelateral density variations are caused by facies varia­tions cannot be ruled out on the basis of gravity dataalone. However, several observations suggest that thisis extremely unlikely. First, the variations occur be­tween two boreholes (WIPP 14 and WIPP 34) slightlymore than 1000 ft apart, and they affect most of boththe Rustler and Dewey Lake Formations. These for­mations were deposited over tens of thousands ofsquare miles in a broad depositional basin (Snider,1966, Figure 34), and both holes are within this basin.C. L. Jones (1954, p 110) notes that while some halitemembers in the Rustler Formation thin reefward, thetwo dolomite members and several silt and sand layersform remarkably persistent stratigraphic markers.For the facies interpretation to be feasible, therewould have to be a very localized depositional anoma­ly within the basin that persisted through the deposi­tion of both formations. Considering the indicationsof halite dissolution within the Rustler Formation(Snyder, this volume, Sec 2.4.3; Ferrall and Gibbons,1980; Bachman, 1980) the karst interpretation ismuch simpler.

81

The gravity survey was unsuccessful in its originalobjective of delineating the DZ structures. Both thebroad anomalies apparently originating from lateralvelocity/density variations and the sharp negativeanomalies, interpreted by L. Barrows as resultingfrom karst processes, are sufficient to mask anomaliesoriginating within DZ structures.

3.4 Detailed CoreDescription

The primary working tools of a structural geolo­gist are observable folds, lineations, and their orienta­tions. These structures often have a distinctive style,characterized by properties such as fold shape (isocli­nal or open) and asymmetry or by observations ofwhat mineral(s) or structure(s) defines a lineation.Such information often allows separation of the ob­served deformation into different events character­ized by distinctive structural styles. The events mayrepresent different episodes separated in space and/ortime, or the separate events may signify discontinuouschanges in mechanism during a continuum of onedeformational episode. As these observations and di­visions are made over an area, one may begin toconstruct an image of deformational events in timeand space. Such an areal description has been done forthe WIPP site and is reported here.

3.4. 1 Multiple Fold andDeformation Textures

Several generations of folds and deformation areobserved as follows (see Figures 3-26 through 3-29):

" Isoclinal folds with opposite sense of asymme­try, i.e., Z- vs S-type folds instead of youngerfolds. These isoclinal folds may represent a sedi­mentation or resedimentation accompanying aslumping event (Parea and Ricci-Lucchi, 1972;Ricci-Lucci, 1973).

" Open asymmetric folds related to a penetrativedeformation accompanied by an extension joint­ing and vein system that parallels the axial planeof the open folds similar to a zoned crenulationcleavage of Gray (1977).

" Ptygmatic and often disrupted folding of isolat­ed anhydrite laminae in halite.

" Dimensional halite fabric.

82

anhydrite veinsparallel orconvergent tothe axial plane

Figure 3-26. Deformation Styles in Laminated CarbonateAnhydrite Units of the Castile (WIPP 13, 3727 ft)

open fold.

mas.iveanhydrite /l...r. ...~""­

veina & layera.......,~.......

Figure 3-27. Deformation Styles in Laminated CarbonateAnhydrite Units of the Castile (WIPP 13, 3727 ft)

Figure 3-28. Deformation Styles in Laminated CarbonateAnhydrite Units of the Castile (WIPP 13, 3729 ft)

Barty stageisoclinal fold

Figure 3-29. Deformation Styles in Laminated CarbonateAnhydrite Units of the Castile (WIPP 13, 3733 ft)

The fold styles and opposite sense of asymmetryidentify the early sedimentary or metasedimentarydeformation relative to a later stage, possibly tectonicstructures. The opposite sense of asymmetry and layerconfinement of the sedimentary structures imply thatthese structures could not have formed during thetectonic deformation. The tectonic structures arecharacterized by asymmetric folds, necking, or pullingapart of competent calcite laminae and veining paral­lel to the axial plane of folds developed during thisstage. Structures suggest the presence of a fluid ac­companied by solution and redeposition during defor­mation. Such evidence is the axial plane veining,pressure solution in nodes of folds, and shadow zonehalite in pull-apart structures. The extension frac­tures as described by Kirkland and Anderson (1970)and Anderson and Powers (1978) are not a purelyfracture phenomenon, as attested by bounding organ­ic-rich laminae extending part way across the pull­apart structure. This structure developed in responseto extension and associated pressure solution.

Other solution-related features accompanying de­formation are exhibited at the 2733-ft level of ERDA6, a section described as a fault zone by Jones (1981).However, current reexamination indicates that thissection is a major halite vein that encroached alongthe foliation planes of the Castile. This vein enlargedby dissolution of the country rock and incorporatedclay-band residues that are relics of the Castile folia­tion.

3.4.2 Petrographic Description:Deformation Mechanism

IntroductionMicrostructures are revealed through petrograph­

ic description, and clues for determining the deforma­tion process come through such study. These impor­tant clues are often the destruction of preexistinggrains and the formation of new grains related to thekinematic history of the rock. Observations like theseallow an initial determination of relative ages fordeformation events. Also, intergranular textures andinternal textures of individual grains often revealsome information such as saturated grain boundariesor twinning, which helps to pinpoint which deforma­tion mechanisms have occurred.

83

Several distinct processes manifest deformationmechanisms in rocks:

• Grain reorientation• Grain breakage• Plastic deformation• Dissolution - precipitation reactions

Each of these processes will characterize the deforma­tion mechanism at different stresses, strain rates, andtemperatures. To understand the deformation of thedisturbed zone, we need to determine which mecha­nisms were active within the halite and anhydriteunits.

Microscopic TexturesIn the experiment conducted by Muller et al

(1981), the anhydrites that were deformed at roomtemperatures exhibited undulatory extinction. Athigher temperatures, twinning became the dominantmicrostructural feature in weakly deformed anhy­drites. As deformation increased, lattice reorientationoccurred by shear twinning and twin boundary migra­tion. Dynamic recrystallization is evident at highertemperatures in the region of steady-state flow. Su­tured grain boundaries suggest that grain-boundarymigration occurred instead of nucleation and growthor subgrain rotation.

WIPP 13 Microscopic TexturesThin sections from WIPP 13 at depths of 3667.4,

3727.7, and 3733.5 ft were made for petrographicstudies (Figures 3-31 and 3-32).

These samples exhibit microboudinage of calcitelaminae (see Figure 3-32) as expected because of therelative strengths of calcite and anhydrite. However,the microboudinaged calcite laminae do not formperfect jigsaw puzzles. Pull-apart boundaries are dif­fuse or embayed, suggesting some dissolution of theboundary.

Anhydrite occurs in the fine-grained matrix mosa­ic or as coarser grained neoblasts (new relative growthgrains; see Figures 3-31 and 3-32). The neoblastsexhibit pressure shadow growth in the pull-aparts ofthe microboudinaged calcite laminae. The deeperspecimens, 3727.7 ft and 3733.5 ft, display nucleationof anyhydrite on the pull-apart surface and elongationof the anhydrite neoblasts. Some elongated neoblasts(see Figure 3-31) contain helicitic inclusion trainssuggesting synkinematic growth and rotation (Ver~non, 1975). Such rotated neoblasts are similar topropellor chloritoids in the Alps (Zwart and Calon,1977). Despite the suggestion of the synkinematicgrowth and rotation, the anhydrites lack twinning andundulatory extinction.

5000

100

50

1000

500

(barsl

500400100

--- -- Halite- . - •- Riburg Anhydrite-Wandflue Anhydrite

'. - • - • - Solnhofen U.nestone

I 1~o

\ '... 10-10\ \

\ ",10-14 \.\ . \

, 1"0-'4 \'~ \ ~

~0-14 ..........\.. \, ....... .," '. ........... .,

... \ .... <Ill,1....__'..;:'T-...;·~--,,....1........J.,...:~_.;,...__.:·,.......Jo

4

Experimental StudiesMuller and Briegel (1978) and Muller et al (1981)

investigated the rheological behavior and deformationof natural anhydrite. Generally, the strength of anhy­drite was found to be between that of halite and fine­grained limestones (see Figure 3-30). Intracrystallineglide and twinning were the major deformation mech­anisms (Muller et aI, 1981). The transition from work­hardening to steady-state flow at low stresses corre­lates with the onset of dynamic recrystallization bygrain-boundary migration. For geologically reason­able strain rates (10-10/S to lO-14/s), the temperaturethreshold for drastic strength reduction in anhydriteis placed between 180°C and 200°C.

200

Temperature (OC)

Figure 3-30. Synoptic Diagram of Log Stress vs Tempera­ture Illustrates the Relative Strength at 10% Strain ofHalite, Anhydrite, and Limestone Expected Under Geologi­cal Strain Rates

f..S 3

I!!:"'02

i

84

c £

; J'~ '."",., ;i-(-

:e

~- .~.~

~.'.,.. /'''.1: '~'.,.,.~ . _ .....'.. .. . ---

'''' :.~'. '" ''''. --'"1·~ J. __n-,.... "'.:' "INA ... ~".,.. " -" ,

........~..... .Mil. '....J~ ...

...s '''-~~.~ ... f,....~.-.. '''~'-.. ~> ~" ". :-- . ;1"~.. ... '...........(';L"f'.~.'.~f:"~ .:: \,.".1:7' . ,:

'At,fttjll' .............

''l'',~'''''~~/ '::: ",-:"~,,.-"_fit ...···--- .. .' -lI! C~d . .. ~" ....

d~/i.~..,.~. .., " ' ..~t ",· ..,...,... .

1mm I, NA-rotated anhydrite neoblast

MA-matri x anhydrite

CC -laminae of calcite, clay & opaque mineral

- 5- trace of folded S surface

Figure 3-31. Rotated Anhydrite Neoblast; Synkinematic Growth Suggested by Helicitic Calcite-Opaque Mineral Inclusion

00U1

•• .1m..m.. ' NA-neoblast anhydrite

MA -matrix anhydrite & clay

Mc-microboudinaged calcite laminae

Figure 3-32. Pressure Shadow Growth of Anhydrite Neoblast; Microboudinage of Calcite Laminae

4. Hypotheses of Origin

4. 1 Gravity FounderingAt least some tectonic stress results from gravity

acting on the earth's complex density structures. Suchstructures may be formed by processes other thanstructural deformation, such as thermal expansion,mineral phase changes, or initial deposition. In thesecases the density structures can be primary sources oftectonic stress. The tectonic stress comes from thetendency of the materials to deform towards a state ofminimum gravitational potential energy in which thedenser material is on the bottom and the lighter is onthe top. Even global plate tectonics probably resultsfrom the gravitational rise of hot low-density materialat the ridges and from complementary sinking of coolhigher-density material at the trenches (Jacoby,1970). Concepts of gravitational tectonics are devel­oped in the texts by DeJong and Scholten (1973) andHans Ramberg (1981) and in articles by Artyushkov(1973), Barrows and Langer (1981), Dennison (1976),Milici (1975), Price (1971, 1973), and VanBemmelen(1976).

The density stratification of the Castile Forma­tion is particularly striking. Figure 3-17 is the com­pensated neutron density log of the Castile Formationat borehole AEC 8. The three massive anhydrite mem­bers with a density of -2.95 g/c3 are separated bylighter halite members. This density structure is grav­itationally unstable. There are tendencies inherent inthe density structure for the massive anhydrite to sinkbelow the less dense halite, and the less dense halite torise above the anhydrite. Whether the deformationoccurs, and its rate, depends on the physical proper­ties of the actual rocks. Ramberg (1981) has extensive­ly investigated systems similar to the density stratifi­cation in the northern Delaware Basin. Part of thiswork involves the mechanical analysis of the gravita­tional deformation of a sequence of initially horizon­tal, viscous layers of varying densities, thicknesses,and viscosities.

In the Appendix to this report, these analytictechniques are applied to the regional density stratifi­cation of the WIPP site. The analysis is consistentwith high-amplitude closely spaced folds in the Cas­tile Formation; low-amplitude broad folds at the topof the Salado Formation; and little or no deformationat the ground surface, the repository horizon, and inthe Delaware Mountain Group. If effective viscositiesare assumed that are typical of those used to modelsalt diapirs, (1018 P), the rate of deformation is0.05 em/yr.

Another Dart of Ramberg's work involves centri­fuge modeling of gravitational deformations. Suchmodels are significant to tectonic studies because ifthey are properly scaled they pass through a structuralevolution identical to that of their geologic prototype.In the Appendix, the scaling ratios of centrifuge mod­els or orogenic belts are redefined so that modelsrepresent gravitational foundering of massive anhy­drite through less dense halite. The models provide anappreciation of the structural complexity resultingfrom the gravitational process. There is an inferencethat the required duration for development of the DZis about 700 000 yr.

The apparent stability of undeformed areas is lesswell understood. A possible explanation lies in thedependency of gravitational shear stress on the ampli­tude of existing deformations. As noted in the Appen­dix, the shear stress at the deforming interface be­tween two gravitationally unstable layers is propor­tional to the amplitude of the deformation. Nodeformation would occur if the evaporites possess afinite yield strength that is higher than the gravita­tional stress inherent in naturally occurring low­amplitude structures. Secondary processes, such asexternal tectonic faulting, could produce initial struc­tures whose gravitational tectonic stress is above thisyield strength. The DZ deformation would then grow

87

outward from this area. An indication of such initiat­ing structures is the high on the top of Anhydrite I atWIPP II.

An alternate interpretation for localization ofcomplex structures within undeformed areas is sug­gested by the effect of water on evaporite rheology.Wenkert (1979) noted that the strain rate in Iraniansalt glaciers was very much larger than predicted bypreviously investigated halite flow mechanisms. Heattributed this to an interstitial solution of saltwater.D. Borns (Sec 3.4) suggested that low-temperaturepressure solution facilitated by water may be a defor­mation mechanism in the DZ. Anomalously high wa­ter content may be supported by the brine pocketsencountered in disturbed structures in the northernDelaware Basin. In this interpretation, the complexstructures form in areas of anomalously high watercontent or anomalously distributed water because thewater facilitates grain boundary pressure solution.The brine pockets form when intergranular watermigrates into low stress regions in the developingstructures.

The gravitational foundering mechanism is con­sistent with available observations. The mathematicalanalysis and centrifuge models presented in the Ap­pendix suggest that DZ structures are the expectedconsequence of existing density stratification, alongwith reasonable estimates of evaporite viscosity.

4.2 DissolutionMechanisms

Dissolution of salt in the Castile Formation hasbeen proposed as the origin of the structures near thesite and in the northern Delaware Basin. It has beenproposed in two somewhat different forms: (1) disso­lution near the top of the evaporite section may havecaused collapse and fill, and the reduced local densitypermitted structural deformation to start (Andersonand Powers, 1978); (2) dissolution in the Castile (orperhaps lower Salado) removed halite and resulted indeformation of overlying and surrounding beds, as inthe DZ (e.g., Anderson, 1981). Neither form of themechanism recognizes a potential role for intergranu­lar fluids in changing the mechanical behavior of therock. The association of deformation and brine wasmade, but the various discussions (e.g., Chaturvedi,1980) indicate that brine is considered to have beentrapped by the structure without playing an active

88

role in its deformation. We will discuss our presentconcept of the role of fluids in deformation in Sec 6.4.

Lambert (1983) has summarized the evidence fordissolution as the origin of various features around thenorthern Delaware Basin. In this discussion, we con­sider only features within the Delaware Basin thatappear nominally similar to the DZ. We do not discusssingle borehole anomalies in salt thickness nor variousmanifestations of dissolution such as "breccia pipes,"karst mounds or domes, and the like. Of interest hereare the DZ, the deformation adjacent to the reefmargin, structures of the Poker Lake field, and thestructure southwest of the site at the James Ranchfield. These features are amenable to some analyses ofthe role of dissolution.

How can we examine the effects of dissolution onsuch deformation? If near-surface dissolution has ini­tiated deformation, the remains of collapse-and-fillstructures should be prominent in the region of defor­mation. The Bilbrey Basin and San Simon Swale areashave fill of probable lower density sediments. (Thedifferentiation between erosion and shallow dissolu­tion is unimportant here.) This is consistent with thegeneral statement of the hypothesis (Anderson andPowers, 1978). The DZ near the WIPP site has somedissolution indications of such shallow densitychanges, according to high-precision gravity data.However, as shown by the Rustler isopach (Figure2-8), the more general pattern of shallow dissolution isroughly north-south and exhibits very little correla­tion with the boundary of the DZ. Equivalent detail isnot available for the James Ranch anticline and thePoker Lake structures.

Dissolution within the evaporite beds, particular­ly the Castile and lower Salado, is supposed to have amore direct effect on deformation by removing salt.Two effects, apart from deformation, are probable: theformation of dissolution residues and the net loss ofsediment from the section. Recrystallized halite hasbeen taken as prima facie evidence of dissolution (e.g.,Anderson, 1981), but for the DZ there is no suchunique relationship to the large deformation features.Likewise, the brine reservoirs in deformed evaporiteshave been associated with dissolution by Anderson(1981). He also invokes a connection in the past (andpossibly in the future) with the underlying fluid­bearing unit of the Delaware Mountain Group. We donot believe there is a major credible association be­tween the Delaware Mountain Group, regional disso­lution, and brine reservoirs, because of the chemistry(Lambert, 1978 and 1983) and the small volume ofbrines.

4.3 Gravity SlidesGravity tectonics is a highly scale-dependent con­

cept. Large-scale phenomena such as :'lithosphericplate interactions can be attributed to gravitationalprocesses. However, gravity tectonics in the restrictedsense is the process where at crustal levels a regionallyintegrated tectonic system has lost potential gravita­tive energy during deformation (De Jong and Schol­ten, 1973). Such reduction of free potential energy canbe accomplished either through vertical displacement(diapirism), or by gravity slides that are dominantlylateral displacements. Diapirism, or halokinesis, isdiscussed in the Appendix. Gravity slides are the focusof this section.

Two processes of gravity slides can be envisagedfor the WIPP site deformation. One is the initiation ofgravity slides by basin tilting. The other is slidesproduced by density contrasts within interfingeringanhydrite-halite sequences at the reef margins. Prob­lems with such processes are the relative timing ofdeformation and tilting, and the question whether theangle of tilting or the strength of density inversion isadequate to initiate deformation.

Another objection would be the lack of detach­ment or decollement surfaces for the gravity slides.However, a zone of low viscosity such as the halitelayers in the Castile can facilitate mass movementrather than discrete surfaces (De Jong and Scholten,1973). The problems of timing and adequate energy toinitiate deformation are discussed individually below.

The important question rises whether deforma­tion occurred coeval to or after basin tilting. Deforma­tion is related by Kirkland and Anderson (1972) andAnderson and Powers (1978) to tilting because of theparallelism of fold hinges and strike of the basinduring tilting. The timing of basin uplift and tilting isplaced as follows by King (1948), Hayes (1964), andAnderson (1981), respectively:

• Minor uplift and exhumation in the early Ceno­zoic

• Earliest identifiable phase of uplift, probably inthe Miocene to early Pliocene

• Main phase of uplift, in the late Pliocene to earlyPleistocene

However, Jones (1981) places the ERDA 6 anticline aspre-Ogallala Formation and therefore as pre-Miocene.Thus, some of the site deformation occurred before

the suggested main stage of tilting. The interrelation­ship of basin flexure and deformation may be critical,but it is not clear and possibly could be only a covari-ance. .

Regarding the question of adequate strength,Wenkert (1979) calculated the shear stresses at thebase of salt glaciers. Under the conditions of tempera­ture, strain rate, and calculated stress for the glaciersunder consideration, Wenkert concluded that defor­mation did not occur through either of the mecha­nisms of Nabarro-Herring creep or dislocation climb.However, he proposed a deformation mechanism ofion diffusion through an interstitial saltwater solutionfrom intercrystalline boundaries of high stress tothose of low stress. This process is equivalent topressure-solution mechanisms discussed elsewhere inthis report. Similar calculations can be made for theWIPP site and result in a shear stress an order ofmagnitude higher than Wenkert's calculation for thesalt glacier (2 x 105 Pa compared to 3 x 104 Pa). Still, itis not clear, based on experimental and theoreticalconditions, whether the slope of the Delaware Basin isadequate to initiate salt movement through the mech­anisms of creep and/or dislocation climb. Again, thepresence of an intergranular fluid may lower thethresholds of deformation. Much work remains to bedone on halite and anhydrite petrofabrics from theWIPP site to determine the active deformation mech­anisms.

In summary, the relative timing of deformationand basin tilting is critically important in ascertainingwhether gravity sliding was or is a mechanism in sitedeformation. Some field evidence in the case of theERDA 6 anticline suggests that deformation occurredbefore the main stage of basin tilting. However, wemust be wary not to preclude earlier basin tilting.Such tilting, although minor, could have been relatedto the Nevada-Sevier and Laramide orogenies thataffected the Cordillera during the Mesozoic. It isharder to evaluate the potential for gravity slides thatare related to the density contrast and inherent meta­stability at the reef margins. Again, under the lowtemperatures of the Delaware Basin, the stresses pro­duced by such density contrasts may not be adequateto initiate deformation without weakening by inter­granular fluids, or unless deformation occurred in thePermian before consolidation. Kirkland and Ander­son (1970) argue ,that several meters of laminated

89

anhydrite and carbonate involved in the folding con­tain 10 000 yr of varves. Their argument continues inthat it is unlikely that the sediments remained uncon­solidated over such periods of time; therefore, defor­mation occurred after consolidation. Also, thesyndepositional folding described in this report islayer-bound; later structures involve multiple layers.Hence, most of the deformation is inferred as youngerthan the Permian deposition of units in the basin.Reef margin gravity slides cannot explain the defor­mation that occurs in the center of the basin. Inconclusion, lateral displacement gravity tectonics orgravity slides may be attractive processes for the sitedeformation, but there are serious mechanistic andtiming problems.

heterogeneous pockets of gypsum. Such pockets couldbe relics of sedimentary processes. However, littleevidence exists to suggest such a distribution of gyp­sum. Also, regardless of the temperature reduction ofthe dehydration equilibrium (Eq (1» caused by lowerH20 activity, the units in question may not have beenburied nor heated sufficiently to bring about dehydra­tion if gypsum was present. Gypsum pseudomorphsafter anhydrite are rarely observed, and some struc­tures such as bladed anhydrite clusters in Anhydrite Iof DOE 1 and WIPP 13 suggest primary anhydritegrowth. Therefore most deposition need not have beengypsum, and the observable evidence indicates thatanhydrite is primary. Hence, the gypsum dehydrationhypothesis for the WIPP site is at best conjecture.

4.4 Gypsum DehydrationKirkland and Anderson (1970) discuss the histori­

cal development of the gypsum dehydration hypothe­sis. From the experimental standpoint, Heard andRubey (1966) proposed that deformation in evaporitesequences and movement of thrust sheets with basalzones of evaporite were facilitated by the dehydrationof gypsum.

This hydraulic weakening was suggested by anobserved tenfold decrease in strength during Heardand Rubey's experiments. The process envisaged isthe increase in fluid pressure within relatively imper­meable rock sequences caused by thermal decomposi­tion of hydrous minerals such as gypsum. For gypsumin the experimental runs, this decomposition andassociated strength decrease occurred over the tem­perature range of 100°C to 150°C. In the Gulf Coast,such temperature thresholds would be encountered atdepths of 2500 to 6000 ft. However, Heard and Rubeyhave not considered the effects of lowered H 20 activi­ty in associated brines. Such lowered activity as ex­pected in nature would lower the equilibrium tem­perature of reaction, Eq (1).

In the case of the deformed evaporites of theWIPP site, the dehydration hypothesis would requirethat areas within the lower evaporites such as theCowden and the anhydrites of the Castile contain

Ca S04 . 2H20 = Ca S04 + 2H20gypsum anhydrite water

(1)

4.5 DepositionalProcesses

Several depositional processes have been suggest­ed as mechanisms for the deformation observed in theDelaware Basin:

• Penecontemporaneous folding• Resedimentation• Slump blocks off reef margins• Sedimentation on inclined surfaces

Penecontemporaneous folding or movement associat­ed with deposition is commonly diapiric (Gale, 1980).Accompanying this deformation, reexposure of therising units occurs. Downbuilding has been invoked asthe possible syndepositional process in the DelawareBasin (Snider, 1966); however, reexposure features arerarely observed in the Castile evaporites. Also, thethickness of varied Castile sequences that are affectedby deformation requires thousands of years for accu­mulation. Such time spans would imply consolidationbefore deformation (Kirkland and Anderson, 1970).

Resedimentation of evaporite sequences has beenobserved in the Mediterranean region (Parea andRicci-Lucchi, 1972). Such units occur as slumpedbodies overlying or replacing normal evaporites in theslope area. Also, gypsum turbidites are associated withsandstone turbidites. These resedimented evaporitesare attributed to overloading and to density instabil­ity of newly deposited beds. Resedimentation is char­acterized by chaotic levels in the sequence; such levels

90

exhibit slumping and mud flows that are mixed withother clastic material. In contrast, the units of theWIPP area show little chaotic or clastic structures.Also, Kirkland and Anderson's (1970) argument thatthe deformed units were consolidated at the time ofthe deformation applies to resedimentation.

Slump blocks off reef margins' into the basinwould result in olistoliths and blocky units as de­scribed above in Parea and Ricci-Lucchi's (1973) re­sedimented evaporites. Again, such phenomena arenot identified in the WIPP area.

If the structures observed were the result of sedi­mentation on inclined surfaces, then the steepness ofsuch surfaces, up to 700 (e.g., Anderson and Powers,1978), would be unreasonable. Again, the inferredconsolidation of units argues against such sedimenta­tion on inclined surfaces.

Overall, depositional processes are inconsistentwith most of the deformation observed. However,some overturned tight folds as seen in core slabs,inclined unconformities, and scour marks suggestsome syndepositional deformation. Still, such struc­tures are only a minor component of deformation nearthe site.

91-92

5. Discussion

5.1 Data BaseThe WIPP site is one of the most intensely stud­

ied bedded evaporite areas in the world. Regionalgeology is known both from outcrops along the west­ern side of the Delaware Basin and from subsurfacepetroleum exploration. Petroleum exploration hasprovided regional gravity, aeromagnetic, and reflec­tion seismic surveys, along with many boreholes. Inaddition, the potash mining industry has cored manyholes through the McNutt ore zone and has ultimatelyestablished the geologic structure for the mined-outareas by virtue of several thousand miles of drift.

Local WIPP site characterization has includedgeologic, hydraulic, geochemical, petrofabric, and geo­physical studies. Seventy-eight boreholes have beenspecifically drilled for WIPP investigations, includingnine that penetrate the Castile Formation. The geo­physical site surveys include electrical resistivity, seis­mic refraction and reflection, surface magnetic profil­ing, first-order leveling, and high-precision gravity.The data base on the site is extensive; the primarytask is to interpret these data.

5.2 Geologic StructureBoreholes and seismic reflection surveys have es­

tablished very gentle dips and minimal structure overmost of the site. There is a seismically indicated faultin the southeast quadrant of Zone IV and possiblefaulting along the sides of an elongated syncline in Sec29,30. Both these features appear limited and are nearthe resolution limits of the present seismic data. Off­sets are not expected to exceed 70 to 140 ft (10 to 20 mstwo-way time at 14 000 ft/s), and recent fault scarpsare not present at the surface.

The broad open folds indicated on the seismictime-structure maps may be caused either by geologicstructure or by lateral variations in velocity within theoverlying formations. Gravity data along with uphole

velocity surveys support the existence of lateral varia­tions in velocity and density within the Dewey Lake,Rustler, and Salado Formations.

The principal structural feature near the WIPPsite is the DZ. This area was identified as anomalouson seismic reflection data and was later found bydrilling to be a structural complex. The areal extent ofthe DZ includes all or part of Sec 2- 5, 8, 9, 16, 17, andthe northern fifth of Sec 20, 21; it is open to thenortheast. Boreholes in the DZ that penetrate theCastile Formation include ERDA 6, AEC 7, WIPP 11,WIPP 12, WIPP 13, UNION Fed. 1, POGO Fed. 1,and POGO SCL Fed. 1.

The DZ is one of several structural complexes ofthe Castile Formation in the northern Delaware Ba­sin. Most are along the northern periphery of the reef­bounded basin, but some occur as isolated structuressurrounded by undisturbed strata. We have assumeda common origin for these Castile Formation struc­tures. However, in view of the multiple hypothesesproposed and the limited geologic control, it is possi­ble that several processes are involved. Other structur­al features in the northern Delaware Basin are suffi­ciently distinct from the DZ to assume that theyresulted from different processes. These include the"breccia pipes," Bell Lake and Slick Sinks, the igneousdike, and San Simon Swale.

Both borehole and seismic data indicate that themost intensive DZ deformation is in the Castile andlowermost Salado Formations. Some structure in theunderlying Delaware Mountain Group is indicated bythe "high" on the top of Anhydrite I at WIPP 11 andthe interpreted faulting in Sec 1, T23S, R30E. Howev­er, the vertical displacements are much less than arecommonly found within the Castile Formation. Theshallower structure at the top of the Salado Formationis characterized by broad, low-amplitude, open folds.This includes a syncline over the DZ just north of thesite and broad anticlines at ERDA 6 and in Sec 1, 2 of

93

T23S, R30E. Above the Salado Formation it is diffi­cult to distinguish geologic structure from the effectsof near-surface dissolution.

It is difficult to establish the true form of thegeologic structures. The "bull's eye" character of theborehole-controlled structural contour maps, alongwith the complex structure observed in the cores,strongly implies that present borehole control is insuf­ficient to adequately resolve the structures. The seis­mic lines support the presence of a more complexstructure than that appearing on the borehole­controlled structural contour maps. For example, line77X7, shotpoints 117 through 130, shows strata dip­ping toward ERDA 6; the borehole-controlled mapindicates a regular anticline.

5.3 Age of DeformationThe age ofthe DZ is equivocal and likely to remain

so. An angular unconformity has been identified at1744 ft near the middle of the Salado Formation inERDA 6, indicating that at least some deformation isPermian. However, the unconformity is localizedwithin a normal section, and it may not be related tothe DZ structures. Anderson and Powers (1978) notedthat the extension fractures in Anhydrite II at ERDA6 cut across microfolds in the anhydrite/calcite lami­nations. Kirkland and Anderson (1970) have arguedfrom microfold orientation and relation to larger scalefolds that the microfolds developed during Cenozoictilting of the Delaware Basin. The implication is thatthe DZ deformation at least partially postdates theregional tilting. However, the microfolds are irregular­ly distributed in the basin, and those examined byKirkland and Anderson may be genetically unrelatedto those found at ERDA 6. At ERDA 6 the microfold­ing and extension fractures may have both been pro­duced by multicomponent deformation during a sin­gle deformation event.

Another approach to establishing the age of theDZ is to apply the axiom that the deformation mustpredate deposition of the oldest undeformed strata.The DZ would then have largely formed before middleSalado time. Subsequent deformation such as ob­served at the top of the Salado Formation is limited tobroad, low-amplitude, open folds. In an attempt toestablish a minimum age, Jones (1981, p 18) notes thatuplifted and arched Triassic rocks near the ERDA 6

94

borehole are truncated by the flat-lying, undeformedPliocene Ogallala Formation. He interpreted this as 0an indication that salt movement was complete before •deposition of the Ogallala Formation. However, hedoes not explain either how the Triassic structurerelates to the deeper DZ or how it is distinguishedfrom near-surface dissolution effects.

The problem with this approach is the inherentassumption that deformations persist upward intooverlying formations. Most of the salt deformationprocesses considered in this report (gravity founder­ing, dissolution, gravity sliding, and gypsum dehydra­tion) could conceivably occur at depth with littleeffect on the free surface.

Finally, some age indications might be inferredfrom the fundamental mechanics of the deformationprocess. Depositional variations would, of course, bePermian. Gypsum dehydration could also reasonablybe assumed a penecontemporaneous process. If thecalcium sulfate was originally precipitated from sea­water as gypsum, then in a normal geothermal gradi­ent the gypsum is stable at depths of above 800 m(2625 ft). Below these depths anhydrite is the stableform so that gypsum-anhydrite conversion shouldhave occurred before complete deposition of thePermo-Triassic Dewey Lake Formation. Further dis­cussion of the gypsum-anhydrite system is available inBraitsch (1971, pp 37-40). The gypsum dehydrationmechanism would presently be inactive because nogypsum is present in the Castile Formation within theDZ.

The gravity gliding or decollement mechanism isreasonably attributed to regional tilting of the basin,but relating the timing of the DZ deformation to thattiming of regional tilting is complicated. The difficul­ty is that gravity gliding is the combined result of allthe sloping density interfaces in the system, includingthe free surface. The free surface may be an erosionalproduct whose slope is not directly related to thedeeper structure. If we ignore this complication, thenthe DZ deformation should be synchronous with theregional tilting. In his discussion of deep-seated saltdissolution, Anderson (1981, p 144) discusses the tim­ing of the regional tilting:

The advanced stage of dissolution in the Dela­ware Basin developed after a relatively recenthistory of uplift and exhumation. Accordingto King (1948, pp 120-122) minor uplift mayhave taken place in early Cenozoic time, but

the earliest identifiable phase was probably inMiocene to early Pliocene time. The mainphase of the uplift occurred in late Pliocene toearly Pleistocene time (King, 1948; Hayes,1964), or perhaps 2-6 m.y. ago. A more recentstudy since those of King and Hayes hasidentified the Salt Basin grabens, the Guada­lupe Mountains, and the Delaware Basin up­lift as features marginal to, and associatedwith, the Rio Grande rift (Seager and Morgan,1979). The development of the rift in southernNew Mexico culminated about 4 m.y. ago.

A Plio-Pleistocene age can thus be inferred for thegravity gliding mechanism. However, Bachman (1980)argues for an earlier stage of uplift based on anangular unconformity of Cretaceous sediments overTriassic and Permian units; and Chap 2 argues for apre-Pliocene cutoff.

Age implications of the salt-dissolution mecha­nism are less certain. According to Anderson (1981)some salt was dissolved from the top of the CastileFormation before deposition of the Salado Formation;from the top of the Salado before deposition of theRustler Formation; after the Permian and before theCretaceous; and after the late Cenozoic uplift, tilting,and erosion of the Delaware Basin. A salt dissolution/deformation process could then have been operativeover any of these time spans, including the present.

Gravity foundering results from the sinking ofmore dense anhydrite through less dense halite. De­formation stress results from the density inversionand will remain "active" until all the anhydrite hassettled beneath the halite. In this sense gravitationalfoundering should be regarded as a possibly ongoingprocess. In the Appendix it is shown that, for reason­able values of evaporite viscosity, the time requiredfor the DZ deformation (106 yr) is much less than theage of the evaporites (220 x 106 yr). A finite yieldstrength is assumed to retard deformation in theundisturbed normal areas. Conceivably the gravityfoundering mechanism may have operated in the geo­logic past and the yield strength increased later. If so,the deformation would now be inactive. Unless thisconjecture is valid, however, the gravity founderingmechanism implies a potentially active process.

5.4 The Role of PressureSolution

The rarity of twinning and undulatory extinctionin WIPP 13 specimens is in stark contrast to theexperimentally deformed microtextures of Muller etal (1981). Since temperatures of the deformed anhy­drite units probably never exceeded 35°C (Powers etaI, 1978), annealing has not occurred. Therefore, somemechanism has taken place other than cataclasis,dynamic recrystallization, or dislocation glide. Theoccurrence of oriented fabrics and stress (pressure)shadows suggests a process of pressure solution(Elliot, 1973; DeBoer, 1977; Robin, 1978).

Pressure solution is the tendency for crystals todissolve in places where the stress component normalto the crystal face is high, and to precipitate simulta­neously in places where this stress component is low(DeBoer, 1977). Confusion can stem from the applica­tion of pressure solution to both sedimentary (diage­netic) and tectonic processes. Diagenetic pressure so­lution occurs in response to gravitational load;tectonic pressure solution occurs in response to weaktectonic/compressive loads (Durney, 1972). Theremay be gradations between the two, but the WIPP 13deformation is an example of tectonic pressure solu­tion.

Stress and permanent strain compete in control­ling pressure solution (Bosworth, 1981). The role ofstress is the migration of chemical components downchemical potential gradients that are functions ofstress gradients (Robin, 1978). Strain effects are thevariation in dislocation density and in turn the solubi­lity product of the host grain (Bosworth, 1981).

By drawing the analogy to quartz where pressuresolution is the dominant deformation mechanism atlow metamorphic grades (low temperature), one cansuggest that anhydrite (and to some extent halite)may be affected by pressure solution at the low tem­peratures of the Delaware Basin. Elliot (1973) notedthat in halite the presence of impurities such as H20on grain boundaries produces an enormous increase indiffusion rates of grain boundaries. The irregular dis­tribution of such H20 impurities may influence thedistribution of deformation at the WIPP site.

95

5.5 Rates of StructuralDevelopment

Barrows in the Appendix to this report has calcu­lated that the time required for halokinetic deforma­tion should be of the order of 106 ranging from 104 to107 yr. Such a time frame would result in a strain rate3.17 x 10-16 S-1 (1 X 10-8 yr- l ). In comparison, Heard(1972) uses a strain rate of 3 x 10-14

S-1 as representa­tive of the natural deformation of halite in Gulf Coastsalt domes. Carter and Heard (1970) report maximumand minimum strain rates for Gulf Coast salt domes of10-11 and 10-15 s-t, respectively. Wenkert (1979) cal­culated an average strain rate for Iranian salt glaciersas 5 x 10-11 s-t, with a range from 10-8 to 10-13 S-1

(note that 10-8 S-1 is a very rapid geological strainrate). Such data set bounds on the range strain rates tobe considered during site deformation.

In an attempt to scale the differences in deforma­tion parameters between Gulf Coast salt domes andthe WIPP site, we can note the strain rate dependenceon (J and T for the possible mechanisms: dislocationclimb; dislocation creep; Nabarro-Herring and Coblecreep; pressure solution (see Table 5-1). Using thesedependencies, we can calculate the ratios of strain ratefor the Gulf Coast and the WIPP site for each mecha­nism (Table 5-2). However, ratios of strain rates maynot be so easily comparable. One mechanism, pressuresolution, may be active at the WIPP site; anothermechanism such as dislocation climb may be active ina salt dome.

Absolute strain-rate calculations are not as simpleas the comparative ratios above. Table 5-3 showscomparative calculations for pressure solution madeby using the equations of Rutter (1976), and fordislocation climb made by using the parameters ofHerrmann et al (1982). The greatest source of error isthe value used for grain boundary diffusivity since thevalue differs for a solution in a narrow 3-nm grainboundary as compared to empirically determined dif­fusivity in a solution of larger volume. This effect isdue primarily to electro-viscous interactions betweenthe solution and the grain boundary. Despite theuncertainties, we can note an average scaling factor of10-2 to 10-3 between the growth of Gulf Coast saltdomes and structures at the WIPP site. Such factorsmake Barrow's theoretically calculated rate of 3.2 x

96

10-16S-1 for halokinetic structures at the WIPP site

consistent with Heard's salt dome rate of 3 x 10-14S-I.

From Table 5-4, we can consider the rate of structuraldevelopment in the context of the long-term WIPPtime frame of ~ 104 to 2.5 X 105 yr. Within this timeframe, if these processes are active at the WIPP site, astructure the size of the WIPP 12 anticline coulddevelop. The worse case is if a WIPP 12-type structuredeveloped that was centered on the WIPP site. Hence,a WIPP 12-scale anticline could develop within 3 x105 yr; such a structure would not disturb the facilitylevel. Fractures could develop in Castile anhydrites,possibly providing a reservoir for brine accumulation.

However, initiation of a structure centering on thesite would be a random event. Site deformation isconcentrated on the reef margins. Given that thedriving mechanisms for deformation have existed forfrom 30 to 200 my, this areal limitation of deformationargues against both a random process and distribu­tion. Deformation probably develops sequentially.Hence, a more appropriate consideration is the rate atwhich the deformation front progresses toward thesite. A first approximation can be derived from ob­serving an average width of 10 km for the DZ adjacentto the reef. If deformation began with basin tilting at30 my, the belt has grown at a rate of 0.3 mm/yr. Atthis rate, 4.6 my would be required for the deforma­tion front to progress over the site.

Another approach to such rate calculations is toview the development of the boundary of a structuresuch as the WIPP 12 anticline as it forms. The geo­metric layout of this approach is shown in Figure 5-1.As the structure rises, its edge progresses outward.The rate of this progression is in proportion to the rateat which the structure rises times the ratio of the

height to the half-width of the structure (rh = do rv).no

Calculations can be made for the progression of theanticline edge toward the site (see Table 5-5). A timeof 3 x 105 yr is derived for this process, by using arelatively rapid strain rate, 10-15 s-t, of the boundingcalculations. The result probably varies by an order ofmagnitude either way (104 to 106 yr).

Hence, within 2.5 x 105 yr, the edge of the defor­mation could just progress to the site center. However,this is only the edge, where deformation is minimal (avertical displacement <10 m in Castile anhydrites).

Table 5-1. Strain Rates as Related to the Deformation'Mechanism

Deformation Mechanism*

Pressure Solution

Nabarro-Herring andCoble Creep

Nabarro-Herring Creep

Coble Creep

Dislocation Creep

Dislocation Climb

Source

Rutter (1976)

Raj and Ashby (1970)(processes combined forpolycrystalline material)

Rutter (1976) (applicableto single spherical grain orpolycrystal when

w(Db

)11"- - «1d Dv

Coble (1963) (applicable tosingle spherical grain)*

Rutter (1976)

Weertman (1968)

Strain Rates

exp -HfIRTtau RT

Hr and Dr are used herein for interfacial-fluid,grain-boundary diffusion to avoid confusion inRutter (1976) in using Hb and Db for both solidand fluid grain boundary diffusion.

• if 11" j (~:) < < 1, then mechanism called

Nabarro-Herring creep

• if 11" j (~:) > > 1, then mechanism called

Coble creep

therefore

derived from combined form above

exp (-HvIRT)tau

RT

exp ( -Hb/RT)taU

RT

. exp ( -H.IRT)t a un

RT

~ a un exp (-QIRT)

RT

where Q is an empirical parameter fromHerrmann et al (1982)

97

Table 5·1. (cont)

Variables

strain rateu deviatoric stressDr diffusion coefficient for an interfacial fluidHr heat of activation for diffusion in an interfacial fluidDb diffusion coefficient for solid grain boundaryHb heat of activation for grain boundary diffusionDv diffusion coefficient for volume diffusionHv heat of activation for volume diffusionR universal gas constantT KV molar volumed grain diameterw width of grain boundary

Approximate

Dr 10-10 cm2 S-l

Db 10-20 cm2 S-l

Dv

10-30 cm2 S-l

Hr 40 kJ mol- 1

H b 110 kJ mol- 1

H v 160 kJ mol- 1

calculated based on Rutter (1976)

*Definitions of mechanisms used herein are from Stocker and Ashby (1973).

Table 5·2. The Effects on Comparative Strain Rates of Thermal and Stress DifferencesBetween WIPP Site and Gulf Coast Salt Domes

Deformation Mechanism

Pressure SolutionNabarro-Herring and Coble CreepDislocation CreepDislocation Climb

Temperature Effects*(~ WIPP/~ Gulf Coast)

Min Max

0.82 0.523 x 10-1 10-4

3 X 10-1 10-4

0.42 10-3

Differential Stress Effects**(~ WIPP/~ Gulf Coast)

Min Max

1.2 X 102 6 X 10-2

1.2 X 102 6 X 10-2

109 10-6

109 10-6

*Conditions used: T WIPP = 308 K

Tl'i~lrc"""t 323 KTl'i:ir Coast = 473 K

Derived by (Heard, 1972) from regional geothermal

gradient; see also Heroy (1968)

98

**Conditions used: lTWIPP = 5 X 10-1 MPalTl'i::Irc"""t = 8 x 10-3 MPa

lTl'i:irc"""t = 8 MPa

derived graphically by

(Heard, 1972, Figure 12)

Table 5-3. Comparative Strain _Rate Calculations for Pressure Solution and DislocationW Climb @ (f = 0.5 MPa and T = 308 K (35°C)

Pressure Solution

where

(J (Rutter, 1976)*

V molar volume solid = 27 cm3 mole-1

Co concentration of solution = 36.9 g/100 c3

Db grain boundary diffusivity = 10-10 cm2 S-1w width of grain boundary = 3 nmp density of solid = 2.16 g/cm3

d grain diameter = 1 cm1.4 x 10-16 s-1

Dislocation Climb

(Herrmann et aI, 1982)

Parameters (A, J.L, n, Q) are from Herrmann et al (1982)

AMn

q

6.7 X 1014 S-1

12.4 GPa4.950.2 kJ mole-1

5.82 x 10-16 S-11.4 X 10-15 S-15.8 X 10-13 s-1

range of Gulf Coastsalt dome temperatures

*Corrected for iteration error in Rutter (1976)

99

Table H. Time Required for Developing WIPP 12·Type Anticlineat 1% Volumetric Strain

100

Strain Rate(~) S-1

5 X 10-11

10-11

3 X 10-14

10-15

3.2 X 10-16

Source

Salt glacier, Wenkert (1979) surface andgroundwater weakened

Closure of mined cavity in salt dome, @25C>C (f = 10 MPa (Serata and Gloyna,1959, in Carter and Heard, 1970)

"Characteristic" geologic strain rate forsalt dome, Heard (1972)

Gulf Coast salt domes, calculated on ba­sis of 1.2 x 108 yr of movement (LowerCretaceous to Recent), € = 3.2 to 4.6,Carter and Heard (1970)

WIPP Site calculation in this report

Time Required forWIPP 12-Type

Structure to Develop(yr)

6.4

32

1.1 X 10-4

3.2 X 105

Table 5·5. Horizontal Rate Calculations forGrowth of an Anticline Edge

If

dorh = - rv

ho '

160mh1------_=_~

fo

do d1

2km

Figure 5-1. Geometry of Boundary Development of aStructure Such as the WIPP 12 Anticline

doho

rv

1000m60m

hI hO

t

17 , rv: ..te of Yenlce' lIl'ow1h

rh: rate of horizontal growth

t:tlme

ho/do =h1/d1d1 =do+(m.tlh1 = ho +(rv. tl... ry/m = ho/do

hI - hO = 60 m

t = 3 X 106 yr @ ~ = 10-16 S-1 .*

Then

rv = 2.0 x 10-4 m/yr

rh = 3.4 x 10-3 m/yr

Therefore,

Time for edge of anticline to reach edge of site:

1000m= 3x105 yr

rh

*assumed most rapid bounding rate

101-102

6. Conclusions

..

This report is a product of several authors withdifferent backgrounds and working philosophies.Such a diverse group will naturally favor differentmechanisms for the origin of site deformation. Still,we basically agree that deformation is largely a resultof salt flowage in Halite I and II. Our ideas diverge onwhat mechanism accounts for this flowage. From thediscussion in Chap 5, we have narrowed the workinghypotheses to three: (1) syndeposition or closely post­depositional deformation that is gravity-driven and inpart initiated by irregular basement topography or byminor basement faulting, (2) gravity foundering, and(3) gravity sliding. However, in the context and thespectrum of objectives of this report, it is not neces­sary to select one favored mechanism of deformation.It is enough merely to examine the effects of eachmechanism and its relevance to site suitability andstability.

6. 1 Constraints on Age ofDeformation

Angular unconformities are observed in the lowerSalado of ERDA 6, and Snyder in Chap 2 describessyndepositional troughs in the Salado. Such observa­tions suggest a Permian component of the deforma­tion. But it must be stressed that it is only a compo­nent. Jones (1981) noted the Ogallala unconformity atERDA 6, which implies a pre-Pliocene deformationstage.

Gravity-driven deformation mechanisms requiredifferent age constraints. The force for gravity foun­dering has been present since deposition in the Perm­ian, and is still a force today. Gravity sliding could beassociated with basin tilting events that have occurredsince deposition through the Mesozoic and Cenozoic;gravity sliding could also be ongoing.

6.2 Syndepositional toClosely PostdepositionalDeformation

The Salado in the DZ is smooth relative to theunderlying Castile. Broad, open, low-amplitude foldscharacterize the Salado in these areas. Also, suchbroad structures are not observed in the stratigraphi­cally higher Dewey Lake Red Beds. These observa­tions led Snyder (in Chap 2) to suggest that the Saladowas deposited onto an already deforming Castile sur­face, and that the deformation was largely completebefore deposition of the Dewey Lake. Meso- and mi­croscopic structures also suggest a syndepositionalstage of deformation. However, these structures arelayer-confined and exhibit opposite senses of asym­metry compared to other sets of structures. Theseother structures (crenulation folds, boudinage, veins,recrystallized matrix) occurred after consolidation, assuggested by the high number of yearly varves in­volved in the structure.

The other crucial question in regard to syndeposi­tional processes is to what extent laws of stratigraphicsuperposition apply in evaporite deformation. Theviscosity contrasts between halite and other units aresuch that the halite may, over geologic time, behavenearly like a fluid relative to the other units. Hence,structures need not be propagated upward throughthe halite units. By means of halokinetic modeling inthe Appendix to this report, Barrows demonstratesthat in one such deformation event, stacked units mayexhibit different wavelengths similar to the observa­tions that stimulated Snyder's suggestion that theSalado was deposited on a deforming surface.

In summary, strong evidence exists for a syndepo­sitional stage of deformation. But equally strong datasuggest that this stage was not the only episode ofdeformation.

103

6.3 Gravity FounderingIn the Appendix to this report, Barrows presents a

theoretical basis for gravity foundering in the Dela­ware Basin. This theory, and the similarity of large DZstructures to those produced by scaled centrifugemodels, are the main evidence for gravity foundering.However, small-scale structures were not observedthat show nonhorizontal flow in the halites, nor wereparasitic folds with reverse drag observed. Hence, noactual observations from the WIPP studies uniquelysupport gravity foundering. Since the driving force forgravity foundering is present throughout the basin, weneed to account for the irregular distribution of defor­mation. In Sections 4.1, 4.3, and 5.4, we have statedthe importance of an intergranular fluid to the initia­tion of deformation. The areal distribution of the DZmay reflect heterogeneities in fluid content within theevaporite section. However, such relationships havenot as yet been empirically demonstrated.

6.4 Gravity SlidingKirkland and Anderson (1970) and Anderson and

Power (1978) observed a parallelism of linear struc­tures in the DZ to basin trends. Such evidence sug­gests that these structures developed as a tectonicresponse to basin tilting. This conclusion is supportedby the observation of horizontal flow indicators inHalite I and II from holes in the DZ as well as in theweakly deformed beds observed in borehole core fromelsewhere in the basin. Still, inconsistencies exist withthe gravity-sliding paradigm. Deformation is distrib­uted around the reef with no obvious down-dip pilingup. Above in Section 6.3, we have restated the rela­tionship of an intergranular fluid and deformation.The evaporite sequence is weakened enough by thefluid, which allows pressure solution to occur, so thatthe units can deform under the low stresses andtemperatures in the DZ. An irregular distribution offluids would limit the distribution of deformation toareas where the fluid content is large enough to facili­tate deformation. Castile units adjacent to the reefconceptually would have higher water content becauseof the availability of reef waters and original sedimen­tation processes. The distribution of deformation fea­tures driven by either gravity foundering or by slidingmay reflect this; however, these areal heterogeneitieshave not yet been demonstrated empirically.

104

The apparent concentration of deformation adja­cent to the reef may also result from (besides heteroge­neous fluid distribution) flow processes in response tobuttressing by the bounding units, irregular topogra­phy of the boundary, and fixed-boundary effects.Holes such as AEC 7, 8, ERDA 10, and DOE 1 inrelatively undeformed areas compared to ERDA 6,WIPP 11, 12, and 13 still exhibit lateral flow texturesin Castile halites. Such observations suggest that saltflowage may exist throughout the basin, but thatcomplex structures (especially in the anhydrite units)are possibly in response to some boundary effect inaddition to the effect of irregular fluid distribution.

6.5 Effects of VariousMechanisms on SiteSuitability and Stability

Distinctions between several mechanisms in geol­ogy can often be equivocal. This is the case with thethree postulated driving mechanisms. Rather thandeciding which mechanism is the dominant one, in thecontext of this report, we can approach the moresignificant question of what the effects of the differentmechanisms are on WIPP site suitability. The pro­cesses can be divided into those now active and thosethat occurred only in the past. Past deformationevents are areally delineated by geophysics and bore­hole mapping. Hence, complicated structures and(possibly) associated brine reservoirs can be avoidedthrough proper location of the site. One must, howev­er, consider whether the processes responsible forcreating existing deformation structures could recurand cause significant deformation at the WIPP site.This concern can be evaluated along with the analysesof ongoing events. Ifdeformation is now active, as maybe true of gravity foundering or gravity sliding, theimportant consideration is the rate of deformation.

Active deformation in the basin is not a randomprocess (see Section 5.5). Deformation structureswould sequentially develop, and the boundary of theDZ would progress' towards the site. As stated above,the rate of this sequential development is the impor­tant consideration. In Section 5.5, the growth of theDZ toward the WIPP site can be calculated at severalrates: 4.2 x 106 yr (using 30 my for initiation ofdeformation and a width of the DZ of 10 km); 3 x 105 yr

..

It

@ ~ = 10- 15 S-1 and 3 x 106 @ ~ = 10-16 S-1 (usingprogression calculation for WIPP 12 anticline edge).There is a remarkable agreement from the variousforms of calculation that reasonable strain rates forthe WIPP site range between 10-15 and 10-16

S-I.

Hence, in predicting the history of the facility through2.5 x 105 yr we can not exclude the possibility that theedge of deformation will reach under the facility nearthe very end of the time frame. This is a bounding butnot absolute calculation. From analogy to the WIPP12 structure, the size of the developing structure willnot displace the repository level. Minor deformationof the Castile units could develop; fractures in Castileanhydrites could provide a reservoir for brine accumu­lation. Fractures that would connect the mid-Saladoand the Castile anhydrites could not develop in halitesat the strain rates involved. Although a brine reservoircould develop at depth, progressing deformationwould not directly jeopardize the facility.

ReferencesR. Y. Anderson, "Correlation and Structural Relation­

ships of the Castile-Salado Evaporite Sequence in ERDA 6,AEC 7, and AEC 8 Boreholes, Eddy and Lea Counties, NewMexico," Open-File Rpt to Sandia (1976).

R. Y. Anderson and D. W. Powers, "Salt Anticlines inCastile-Salado Evaporite Sequence, Northern Delaware Ba­sin, New Mexico," in G. S. Austin, Geology and MineralDeposits of Ochoan Rocks in Delaware Basin and AdjacentAreas, Circular 159 (Santa Fe, NM: New Mexico Bureau ofMines and Mineral Resources, 1978), pp 79-83.

R. Y. Anderson, Deep-Seated Salt Dissolution in theDelaware Basin, Texas and New Mexico, Special Publica­tion 10 (NM Geol Soc, 1981), pp 133-145.

E. V. Artyushkov, "Stresses in the Lithosphere Causedby Crustal Thickness Inhomogeneities," JGR 78:7676-7708(1973).

A. A. Arzi, "Remote Sensing of Subsurface Karst byMicrogravity," in J. S. Tolson and F. L. Doyle, ed, Proc ofthe 12th Inti Congress on Karst Hydrology (Huntsville, AL:UAH Press, 1977), pp 271-272.

G. O. Bachman, "Cenozoic Deposits of SoutheasternNew Mexico and an Outline of the History of EvaporiteDissolution," USGS J of Res 4:2 (1976), pp 135-149.

G. O. Bachman, Regional Geology and Cenozoic Histo­ry of Pecos Region, Southeastern New Mexico, Open-FileReport 80-1099 (USGS, 1980).

L. J. Barrows and C. J. Langer, "Gravitational Potentialas a Source of Earthquake Energy," Tectonophysics 78:237­255 (1981).

Bell and Murphy and Assoc., Inc., Geophysical Investi­gation of the Waste Isolation Pilot Plant, Report for US­DOE (1979).

W. Bosworth, "Strain-Induced Preferential Dissolutionof Halite," Tectonophysics 78:509-525 (1981).

O. Braitsch, Salt Deposits, Their Origin and Composi­tion (Berlin: Springer-Verlag, 1971).

L. Chaturvedi, ed, WIPP Site and Vicinity, GeologicField Trip.:,New Mexico State Environment EvaluationGroup, EEG-7 (1980).

R. L. Cobel, "A Model for Boundary Diffusion Con­trolled Creep in Polycrystalline Materials," J Appl Phys34:1679-1682 (1963)

R. B. DeBoer, "On the Thermodynamics of PressureSolution - Interaction Between Chemical and MechanicalForces," Geochimica et Cosmochimica Acta 41:249-256(1977).

K. A. De Jong and R. Scholten, ed, Gravity and Tecton­ics (New York: John Wiley and Sons, 1973).

J. M. Dennison, "Gravity Tectonic Removal of the BlueRidge Anticlinorium to Form Valley and Ridge Province,"GSA Bull 87:1470-1476 (1976).

US Department of Energy, Waste Isolation Pilot PlantSafety Analysis Report, Vol 1-5 (Albuquerque, NM: USDOE, WIPP Project Office, 1980; rev 1980 and 1981).

M. B. Dobrin, Consultant Report on Seismic Survey ofWIPP Site Conducted by Bechtel in 1979, Report for theUS DOE (1979).

D. W. Durney, "Solution-Transfer, an Important Geo­logical Deformation Mechanism," Nature 235:315-317(1972).

D. Elliot, "Diffusion Flow Laws in MetamorphicRocks," GSA Bull 84:2645-2664 (1973).

A. S. Gale, "Penecontemporaneous Folding, Sedimenta­tion and Erosion in Campanian Chalk Near Portsmouth,England," Sedimentology 27:137-151 (1980).

D. R. Gray, "Morphologic Classification of CrenulationCleavage," J of Geol 85:229-235 (1977).

P. T. Hayes, Geology of the Guadalupe Mountains,New Mexico, USGS Prof Paper 446 (1964).

H. C. Heard, "Steady-State Flow in PolycrystallineHalite at Pressure of 2 Kilobars," in H. C. Heard, I. V. Borg,N. L. Carter, C. B. Raleigh, ed, Flow and Fracture of Rocks(Geophys Monogr Series), Am Geophys Union 76:191-210(1972).

H. C. Heard and W. W. Rubey, "Tectonic Implicationsof Gypsum Dehydration," GSA Bull 77:741-760 (1966).

J. L. Hern, D. W. Powers, and L. J. Barrows, SeismicReflection Data Report, Waste Isolation Pilot Plant(WIPP) Site, Southeastern New Mexico, SAND79-0264,Vol 1 and 2 (Albuquerque, NM: Sandia National Laborato­ries, 1979).

W. B. Heroy, "Thermicity of Salt as a Geologic Func­tion," in R. B. Mattox et aI, eds, Saline Deposits, GSASpecial Papers 88:619-629 (1968).

W. Herrmann, W. R. Wawersik, and S. T. Montgomery,Review of Creep Modeling for Rock Salt (Albuquerque,NM: Sandia National Laboratories, 1982).

105

W. R. Jacoby, "Instability in the Upper Mantle andGlobal Plate Movements," JGR 75 (29):5671 (1970).

C. L. Jones, Geologic Data for Borehole ERDA 6, EddyCounty, New Mexico, Open-File Report 81-468 (USGS,1981).

C. L. Jones, "The Occurrence and Distribution of Potas­sium Minerals in Southeastern New Mexico," NM Geol SocGuidebook 5th Field Conf (1954), pp 107-112.

P. B. King, Geology of the Southern Guadalupe Moun­tains, Texas, USGS Professional Paper 214 (1948).

D. W. Kirkland and R. Y. Anderson, "Microfolding inthe Castile and Todilto Evaporites, Texas and New Mexico,"GSA Bull 81:3259-3282 (1970).

S. J. Lambert, Geochemistry of Delaware BasinGround Waters, Circular 159 (Santa Fe, NM: New MexicoBureau of Mines and Mineral Resources, 1978), pp 32-38.

S. J. Lambert, Dissolution ofEvaporites in and Aroundthe Delaware Basin, Southeast New Mexico and WestTexas, SAND82-0461 (Albuquerque, NM: Sandia NationalLaboratories, February 1983).

G. J. Long and Associates, Report: Interpretation ofGeophysical Data, Los Medaiios and Vicinity, Lea andEddy Counties, New Mexico, Report to Sandia Laborato­ries (1976).

J. M. McGough, DOE-AL, Ltr on WIPP Surface Con­trol Zones, WIPP, JMM 83-0125, February 17, 1983.

R. C. Milici, "Structural Patterns in the Southern Ap­palachians - Evidence for a Gravity Slide Mechanism forAlleghenian Deformation," GSA Bull 86:1316-1320 (1975).

W. H. Muller and U. Briegel, "The Rheological Behav­ior of Polycrystalline Anhydrite," Eclogue Geol Helv 71:397­407 (1978).

W. H. Muller, S. M. Schmid, and U. Briegel, "Deforma­tion Experiments on Anhydrite Rocks of Different GrainSizes: Rheology and Microfabric," Tectonophysics 78:527­543 (1981).

L. L. Nettleton, Gravity and Magnetics in Oil Pro­specting (New York: McGraw-Hill, 1976).

G. Omnes, "High Accuracy Gravity Applied to theDetection of Karstic Cavities," in J. S. Tolson and F. L.Doyle, ed, Proc of the 12th Inti Congress on Karst Hydrolo­gy (Huntsville, AL: UAH Press, 1977), pp 273-284.

G. C. Parea and R. Ricci-Lucchi, "Resedimented Eva­porites in the Periadriatic Trough (Upper Miocene, Italy),"Israel J of Earth-Sciences 21:125-141 (1973).

D. W. Powers, S. J. Lambert, S-E. Shaffer, L. R. Hill,and W. D. Weart, ed, Geological Characterization Report,Waste Isolation Pilot Plant (WIPP) Site, SoutheasternNew Mexico, SAND78-1596, Vol I and II (Albuquerque,NM: Sandia Laboratories, 1978).

R. A. Price, "Gravitational Sliding and the ForelandThrust and Fold Belt of the North American Cordillera ­Discussion," GSA Bull 82:1133-3317 (1971).

R. A. Price, "Large-Scale Gravitational Flow of Supra­crustal Rocks, Southern Canadian Rockies," in De Jong andScholten, ed, Gravity and Tectonics (New York: John Wileyand Sons, 1973), pp 491-502.

106

R. Raj and M. F. Ashby, "On Grain Boundary Slidingand Diffusional Creep," Trans Met Soc AIME 2:113 (1971).

H. Ramberg, Gravity, Deformation and the Earth'sCrust (London: Academic Press, 1981).

F. Ricci-Lucchi, "Resedimented Evaporites: Indicatorsof Slope Instability and Deep-Basin Conditions in Peria­driatic Messenian (Apennines Foredeep, Italy)," in W.Drooger, ed, Messenian Events in the Mediterranean (Am­sterdam: North Holland Publishing Co., 1973), pp 124-141.

P-Y. F. Robin, "Pressure Solution at Grain-to-GrainContacts," Geochem et Cosmochimica Acta 42:1383-1398(1978).

E. H. Rutter, "The Kinetics of Rock Deformation byPressure Solution," Phil Trans Royal Soc London, Ser A283:203-209 (1976).

Sandia National Laboratories and US Geological Sur­vey, Basic Data Report for Drillhole WIPP 11 (WasteIsolation Pilot Plant- WIPP) , SAND79-0272 (Albuquerque,NM: Sandia National Laboratories, 1982).

Sandia National Laboratories and US Geological Sur­vey, Basic Data Report for Drillhole WIPP 33 (WasteIsolation Pilot Plant- WIPP) , SAND80-2011 (Albuquerque,NM: Sandia National Laboratories, 1981).

S. M. Schmid, J. N. Boland, and M. S. Paterson,"Superplastic Flow in Fine-Grained Limestone," Tectono­physics 43:257-291 (1977).

W. R. Seager and P. Morgan, "Rio Grande Rift inSouthern New Mexico, West Texas, and Northern Chihua­hua," in R. E. Riecker, ed, Rio Grande Rift: Tectonics andMagmatism (Washington, DC: AGU, 1979), pp 87-106.

H. I. Snider, "Stratigraphy and Associated Tectonics ofthe Upper Permian Castile-Salado-Rustler Evaporite Com­plex, Delaware Basin, West Texas and Southeast New Mexi­co" (PhD diss, University of New Mexico, 2 parts; p 2 charts,1966).

R. L. Stocker and M. F. Ashby, "On the Rheology of theUpper Mantle," Reviews of Geophysics and Space Physics11:391-426 (1973).

M. Sweeting, Karst Landforms (New York: ColumbiaUniversity Press, 1973).

R. W. VanBemmelen, "Plate Tectonics and the Inunda­tion Model - A Comparison," Tectonophysics 32:145-182(1976).

R. H. Vernon, Metamorphic Processes: Reactions andMicrostructure Development (New York: John Wiley andSons, 1975).

K. H. Waters, Reflection Seismology (New York: JohnWiley & Sons, 1978).

J. Weertman, "Dislocation Climb Theory of Steady­State Creep," Trans Amer Soc Metals 61:681-694 (1968).

D. D. Wenker-t, "The Flow of Salt Glaciers," GeophysRes Ltrs 6:523-526 (1979).

H. J. Zwart and T. J. Calon, "Chloritoid Crystals FromCuraglig; Growth During Flattening or Pushing Aside,"Tectonophysics 43:477-486 (1977).

APPENDIX

Halokinetic* Developmentof the Disturbed Zone

Larry BarrowsSandia National Laboratories

Albuquerque, NM 87185

November 1981(Rev. May 6, 1982)

NoteThis report concerns the mechanical development ofthe disturbed zone and, to a lesser extent, the migra­tion of brine pockets. It is assumed that the reader isfamiliar with the WIPP, its local geology, and thegeneral concepts of gravitational tectonics as devel­oped by Hans Ramberg.

No attempt is made to assess the implications of thismaterial upon suitability of the site. Such assessmentrequires careful consideration of this material alongwith many other factors.

*Halokinesis-autonomous, isostatic movement of salt (Halbouty, 1967)

107

108

AbstractThe disturbed zone was initially identified as an area of chaotic seismic reflectiondata in the northern part of the WIPP site. Boreholes have encountered complexdeformation within the Castile Formation and a broad gentle syncline near thetop of the Salado Formation. The underlying Delaware Mountain Group isrelatively undeformed.

The disturbed zone can be attributed to gravitational foundering of the massiveanhydrite members of the Castile Formation through the less dense halite.Following Ramberg (1981, Chap 7), the process was analyzed as the growth ofsinusoidal waves in a system of viscous layers with differing thicknesses, viscosi­ties, and densities. Analysis of the density stratification of the WIPP site isconsistent with intense short-wavelength deformation within the Castile Forma­tion; less-intense longer wavelength deformation at the Rustler/Salado interface;and little or no deformation at the free surface, repository horizon, and DelawareMountain Group. For effective evaporite viscosities similar to those used inmodeling salt diapirs (1017 to 1019 P), the rate of deformation is 0.5 to 0.005 em/yr.

Centrifuge models of orogenic belts described by Ramberg (1981, Chap 15) werereinterpreted as models of foundering anhydrite layers. The models provide aqualitative appreciation of the structural complexity resulting from gravitationalfoundering.

Other centrifuge models were reinterpreted as representing the rise of brinepockets through the evaporite section. The models suggest upward brine veloci­ties of about 1 em/yr.

Disturbed-zone deformations are the expected result of the WIPP densitystratification, along with reasonable estimates of evaporite viscosity. The stabil­ity of undeformed areas is less understood, but may be due to a finite yieldstrength of the rocks.

~:

ContentsBackground : 111

Observations 111Mechanisms 124

Analysis 124Theory 124WIPP Models 127Rates of Growth 131

Centrifuge Models 133Gravitational Foundering 133Migration of Brine Pockets 137

Discussion 139

109-110

Background

ObservationsThe "disturbed zone" (DZ) is loosely defined as an

area of chaotic seismic reflection data in the northernpart of the WIPP site. Figure A-I is a representativeseismic line (77X5) crossing the DZ. Figures A-2through A-5 are seismic time structure maps, andFigures A-6 through A-8 are seismic isochron maps.These maps show the areal extent of the DZ (FiguresA-2, A-3), the absence of deformation in the underly­ing Delaware Mountain Group (Figures A-4, A-5,A-8), and thickness variations within and on the pe­riphery of the DZ (Figures A-6, A-7). The seismic lines(e.g., Figure A-I) indicated a blocky chaotic structurewith abrupt offsets (faults) and changes in dip be­tween units. Changes in the seismic character indicatevariations in unit thicknesses and/or acoustic proper­ties.

Figures A-9, A-lO, and A-ll show correlationsbetween those boreholes in the DZ that penetrate theCastile Formation. The units in the Salado Formationare convenient markers on the compensated neutrondensity logs. Within the Castile Formation the darkbands are anhydrite; the light areas are halite (a8indicated on the density logs). Brine encounters aremarked by asterisks. The top of the Salado Formationat borehole ERDA 6 is higher than at the flankingboreholes AEC 7, AEC 8, and Union Fed. 1.

Figure A-12 is a top-of-Salado structural contourmap prepared by C. L. Jones, USGS, from boreholecontrol. There is a broad gentle syncline in the area ofthe DZ, but no indication of deformations as intenseas those encountered in the underlying Castile Forma­tion. On Figure A-12 there is a broad gentle anticlineto the southwest of the WIPP site.

Figure A-13 is that portion of seismic line 76SAN2that crosses this area. Locations of this line segmentand the Belco-Federal well (a brine encounter) areindicated on Figure A-12. Seismic line 76SAN2 wasgathered with field parameters typical of petroleumexploration in the region. It does not resolve structuralfeatures within the Castile Formation nearly so well as

the seismic lines gathered with modified field parame­ters (the 77X, 78Y, and 78GG surveys). Despite therelatively poor resolution, the line does indicate struc­ture within the Castile Formation. The seismic char­acter of this structure is similar to that portion of line76SAN3 that is in the DZ north of the site.

There appears to be a spatial relation betweenlow-amplitude folding at the top of the Salado Forma­tion and areas of disturbed structures within theCastile Formation. The top of Salado deformationsinvolves both anticlines and synclines.

The DZ is one of many Castile Formation struc­tural complexes in the northern Delaware Basin. Mostof the others are known only from borehole data,although some are indicated on proprietary petroleumexploration seismic lines. H. Snider (1966) presentsdocumented contour maps prepared from the bore­hole data available at that time. R. Y. Anderson (1978)presents contour maps prepared from the larger dataset then available. In Anderson's report the wellioca­tions are indicated, but the data are not posted on themaps nor tabulated in the text.

Figure A-14 is a structural contour map preparedfrom the posted borehole control. The contoured hori­zon is the top of the first massive halite encounteredbelow the uppermost massive anhydrite in the CastileFormation. Generally (but not always) this is the topof Halite II. The data points were taken from apreliminary contour map by R. Snyder, USGS, andcontoured by L. Barrows. Some of the structuresevident at other stratigraphic horizons are not indicat­ed at this mapped horizon (see Snider, 1966, or Ander­son, 1978). The "bull's-eye" character indicates thatstructures are not resolved by the present boreholecontrol, and it is likely that additional structuresremain undetected in the basin. Nevertheless, FigureA-14 indicates the general distribution of Castile For­mation structural complexes in the northern DelawareBasin.

111

"'It,,.: .....,':IU.: 1&1.-114 ...,....... ClIIlTI:2"I-....e:t: vi_II.....L...: 1211<:.-- "vi'. 1_: 110 H ... ,.".: esc-XID''"''Ie....: 12OCI~T

100r."I"r../.(".o

Illll[a.",

INF0RMATl0N

Las

FIELD

MEDANasLINE X-5

STATI0NS 3-391S0UTHERST NE~ MEXIC0

IT: QlllllII8t ........ ICtct_ J. 1977

"S I - CJlI IV,.I' lie.2'lI-IOO HZltD ~T.

•11&.1"

1210

"IItTlCI ~iIlITI"CRT"",: JIDD fI'T.Ya:i: 6CIGD II".It1C.

II DDU.. T""LEI

2) I'''''' IRI"IIICeWI'tT

J) v' 11 ~'IlTI..

C DB'Tw NUll ..'...-s5> aKlIJIftIkuTIDt

....'. L....'.-I40 "IlLSIIIIIIOI(TI.. T1....0 .. 2tCl ,... C....SIJrilIi

-.-.r,cll

....... ~ 1\ '" lllllG ,.,.... ffltS .

PR0CESSING SEQUENCE

I " I~

l-.u' ......~ 1"'-""

'l !I.-(.... I ...' 0111'....~, "ILlYO.O-J.O.c. ;P.I-.:l loll

~T Dln~ I'''''ICI

.) ll'&.KI'''' ........115-,-10: FIlii'll __ ......11'.. (~Tll

ITlIlIC. 12 ~"'D

(2) ""-IIWM'Ml' OlliTlIl. ~I«:" "ILTliIlIO.O-J.O .C. 2'!5-.c HI

11) DI61'.... .c

U: OIPLIU• './ililto IIll....C.

SNlIIlID 01 ......11 ~ I) 1....................... .. ..

Figure A-1. Representative Seismic Line 73 (77 x 5) Crossing the Disturbed Zone of the WIPP Site

112

•)(

.ID

380

----'>.,~-380

Aft 8

380 /

IIllmll Tlml ItrulturlTI' If Clltlll Fermltll.

...t.., I.t.,,,.. ID .111I.........tI..•••• 1t LL

".,., L." " ..m.""

I--+--\---'·_....",.,-jr-- GG 14

II

IS ...co co Sco ..

..)(...--- .1111'............. II..

0011-- r.lllli. U....................,...... ...... It•• ·., ••" •••t,..,.,•,---- f••II• W,.... .........,...

1--'" 1111. ----l

Figure A-2. Seismic Time Structure - Top of Castile Formation

113

-4.0 _' , .............00

-0,. MII'--I

47.

410

---GGI7--

----_ ....• W'" II

• 1111 14. t•••••• · tl••,. II..I ~ ,.,.. If _,III It,••t.,.f.'t••11•• " 11,1..,

.00 -tS

1t1.1II1. TlIII. 8truotur•Mill. CIIUI. F.rlll.tl ••

..,t••, .,te,.. I, II .'/II.....~•

..t •• ' .... ft.....1.

".,.,.~ I" L"'I .." •••• I".•_m.F.UI

Figure A-3. Seismic Time Structure - Mid-Castile Formation

114

....-.-..M".----4

770

_771- .1111__ ""'.....1_,....1.11.;;17-- I.' ""Inu)------ ...,.

Figure A·4. Seismic Time Structure - Top of Cherry Canyon Formation

.1.11I1, TI... Itr.,t.,.T.p If ell.". c••••• Fe,..tl..

c....., I•••,n'," 11•••••••..••••••10 ft .

".,.,•• '" Ler, , ...... ,..,

115

1240

~

---1- "10

"..

tT--jt----+,.---"~_+- GG 16

IUO

N><

1260

1260

1240

1240

,uo

'270IUO -t6'

..t1I1 ~. two.w.y 11m.,.1 1. II••........ •h... ""'1.''').......1. full

Slllmlo Tlml Struot uriTop If 11111 Sprllli FIrmltllll

...t••, l.t I, '0 011111.......~.t..., 1110 ft 1.

".,.... '1' L'''I .." I.ro•. Ft•. II

Figure A-5. Seismic Time Structure - Top of Bone Spring Formation

116

I

• -" I~

- II

II

la~

/

·-1-'(',13i

•IS

"'-".11.--1

"'1--..........

~

"'" , 'r, , ..,. ."' 11 1" .. If ".,,,.

hl.mic I.echre.Tep C..tll. t. Mid. C..tli.

......r ' ••lr••I: 10 .1111•••••41.,,,,.,,~ '., lI"..." •••. lew. IUD

Figure A-6. Seismic Isochron - Top Castile to Mid-Castile

117

'"

~., M",--C

..ii

"17--

...........

...11....... , , II ...IN.n••, "' ,~ 1141 J,.••••• II,.l,r...04 ••rt ••~I J.... • f ...,1•••, ..

300

280

;

S,I,,,,lo IloohronMid. Clltil. to Ch.rrll C.nllon

a,.t•• , lilt" VII: 10 mllli .• ......4 ~" L.", .." •••• II 1..0

Figure A-7. Seismic Isochron - Mid-Castile to Cherry Canyon

118

4ao

~ I. 4aO

~-~_t-GG\4

,..:j

4aO~

XI ---------1~:__----__;;If_-

418

..,.

t--".lIe---f

l·a"·

...111I....... , •••••• \111I.11Io".I'a .. '.1.....111,.1.11I1.111018 1Il.la"")

StlamlcllochronCharry Cuyon to Bona Spring

Gontour Intlrwll: 10 mUll II."""prop d •• ' l.rry a N••. 1"0

Figure A-B. Seismic Isochron - Cherry Canyon to Bone Spring

119

top S.lado Fonnation

Cowd.n Anhydrite

•

----

/'j '\,~/

!

2000 It ...... -

3000 It ...... -

1000 It •.•.1.-

r"'- - ..i -WIPP II -.- - - - ,r..l

r J

r J

III

rJ

IIIII

I",IL_,

IL~-'-l

,.. ..t1I.I

I

Unit" ,.t1 I·....... ERDA 8 • AEC 7

~V" , AEC 8

Co,L,

L,I"1

IIIIIIII

r JI

,..Jr·

r J

L, r~~____ r-

J ~___ J I mill

~igure A·9. Boreholes in the DZ ThtIle Formation at Penetrate the Cas-

top DelawareMountain Group

Sa. Level-

1000 ft b.•".-

Figure A·11 Saladtions in the DZ (see ~,and Castile Formation Log C 1holes.) , Igure A-9 for horizontal distr'b °t~re a-I u IOn of

e-Anhydrlt.

-------------------2000 It •.• .1. -

1000 ft ...... -

, 000 ft b.•.I. _ ----- tap Dele.....Mountllin Group

Figure A·10 Sal d. . a 0 and C t'lhons in the DZ, (See Fi lS I e Forn~ation Log Correla-of holes.) gure -9 for hOrizontal distribution

120

,(:

--

Fig .

Hudson (a brine encounter)

c

T

.02S

.• ,.,.....:r...

J "& ~ ................ --..

t·t .. tt .......1

11'~........~...­.....,.~..:..

;'/~?; .... , ..~""' .......

.':':::. :.;..~"~:~'t:.~-: •.041

: .. ~•• ,... .' Ir •I"f~ •• •• ~-"" ~ •... ; ...

......: .., ,.1lI " ..• •..1 r••_,.•. '1.. lit,,...,.

~ J ..... ,......., .

.'

N......

Figure A-12. Structure Contour Map on the Top of the Salado Formation (Contour interval: 20 ft)

lep I'P 145I

Figure A-13. The Portion of Seismic Line 76SAN2 Indicated on Figure A-12 (The complex structures in the Castile Formationare loosely coincident with the anticline on the top of the Salado Formation.)

122

()

r-- --------,---------,--

I

Figure A-14. Castile Formation Salt Structure

---"""1-- .--------• I

I

TUS

I-Smi\es-I

Castile fo,,"m~hon SaH- Sn-ucturetor of 1'_ \.,li~ \.elolO ur...",_t ~~e

(mo$l:Iy H,\.klI)

conblu", inttN.I : 100 hcMta: R.SnycAe..... U~E.s.coft\:_... : t..B.Yrow)\o )sw\..Fc~.,'1,'v.1.

123

Documented petrographic studies of cores fromthe DZ are few and limited. R. Y. Anderson (1976)investigated cores from ERDA 6, AEC 7, and AEC 8 toestablish the stratigraphic relations between theseholes. The ERDA 6 core from 2551 to 2733 ft wasexamined in detail. This work showed that the anhy­drite unit in ERDA 6 correlated with the Anhydrite IImember of the Castile Formation. The core in ERDA6 had two solution zones and a corroded zone alsoshowing some evidence of solution activity. Lamina­tion dips were commonly greater than 50° to 60° , andmany of the lamina had parted with the interveningvoids filled with calcium sulfate. These "extensionfractures" postdate microfolding in the Castile Forma­tion. Kirkland and Anderson (1970) have argued fromother data that the microfolding developed during theTertiary.

More recent petrographic studies of core fromborehole WIPP 13 by D. Borns (pers. comm., 1-8-82)have indicated complex polyphase deformation facili­tated by grain boundary solution/deposition. The sug­gestion is made that irregular distribution of watermay influence development of the deformation.

MechanismsThe data base on the DZ is extensive and provides

fairly good control on the structural style. The processand timing of deformation are less well controlled andremain ambiguous. Nevertheless, any feasible processor mechanism must be consistent with the observa­tions, and inconsistent mechanisms can be discarded.All mechanisms that are consistent with the observa­tions should be regarded as feasible. In many practicalproblems, including the dynamic structural evolutionof the DZ, this leads to a variety of possible mecha­nisms. The method of multiple working hypotheses(Chamberlin, 1890) should be used in evaluating thedata and in planning subsequent investigations. Forphenomena as complex as the Castile Formation de­formations, it is possible that more than one mecha­nism is involved. However, in the following, the princi­ple of simplicity or Occam's razor (C. A. Anderson,original date unknown; reprinted 1963) is used toselect one hypothesis as the more likely. This singlehypothesis is expanded in greater detail and accountsfor most, if not all, of the observations.

Mechanisms that have been considered includegravity gliding, dissolution, and external tectonics.Gravity gliding is the eastward movement of shallowerformations in response to the mid-Cenozoic tilting ofthe Delaware Basin. It is difficult to reconcile with theisolated location of some disturbed areas (e.g., PokerLakes), the distribution of the greatest intensity of

124

deformation (i.e., along the northern instead of east­ern basin boundary), lack of identifiable decollementsurfaces, and the intensity of the deformation. Exten­sive salt dissolution seems inconsistent with the ab­sence of massive subsidence of overlying material(some disturbed areas are overlain by anticlines), theabsence of extensive dissolution residues, and the coredescriptions. External tectonics seems inconsistentwith the lack of deformation in the underlying Dela­ware Mountain Group.

At the WIPP site the Castile Formation consistsof three massive anhydrite members separated by twohalite members. The anhydrite members have a meandensity around 2.95 g/c3 and the halite membersaround 2.2 to 2.25 g/c3

• Another possible mechanismfor the disturbed-zone deformations involves gravita­tional foundering of the anhydrite through the lessdense halite. In this mechanism the gravitationallyunstable interfaces between the halite and overlyinganhydrite spontaneously deform into alternating (andcomplementary) halite diapirs and adjacent anhydritesinks. The overlying and underlying strata are vis­cously coupled with the deforming interface. Theentire process is a form of autonomous, isostatic saltdeformation or halokinesis.

The gravity foundering mechanism is further de­veloped in the following sections. It is shown that DZdeformations are the expected, predictable conse­quence of the density stratification of the WIPP sitealong with reasonable estimates of evaporite viscosity.

Analysis

TheoryThe following material is based on the work of

Hans Ramberg at the Institute of Mineralogy andGeology, Uppsala, Sweden. The techniques and theirgeologic justification are best described in the 1981edition of his text Gravity, Deformation, and theEarth's Crust.

Gravitational tectonics is a reasonably well­developed concept in structural geology. In the broad­est sense, deviations from a density-stratified concen­tric global structure represent increases in gravita­tional potential energy. Gravitational tectonic stressdeforms the material towards the state of lower ener­gy. The effective structures can have the form oflateral density variations such as occur oeneath ocean­ic arc-trench subduction zones and spreading centers.In this case the associated gravitational tectonicstresses form the probable driving force behind plate

tectonics. Another effective structure fa a 'density in­version such as develops when a layer of salt is buriedbeneath more dense sediments. In this case, gravita­tional tectonics lead to the development of salt ridges,pillows, domes, and (ultimately) piercement diapirs.

A more general effective structure involves one ormore interfaces across which density decreases withdepth, The total system may involve many layers ofdifferent densities. There may be a net increase indensity with depth through the entire section. Howev­er, if one or more of the interfaces between layers has adensity inversion, the system is gravitationally unsta­ble. Yield strengths, effective viscosities, layer thick­nesses, and densities determine whether and how fastthe potential deformation occurs.

The analysis proceeds from the following assump­tions:

1. The model consists of a sequence of layers in auniform gravity field. The layers extend toinfinity in the two horizontal directions. Thick­ness, density, and viscosity are constant withineach layer.

2. Materials behave as isotropic Newtonian (lin­ear) viscous fluids. The yield strength is negli­gible.

3. Inertia is insignificant.4. Interfaces between layers are continuous. This

implies continuity of displacements and stress­es.

5. The deformations are sinusoidal waves affect­ing the entire depth of the model. The ampli­tude/wavelength ratio is small and the defor­mations of all interfaces are in phase.

As an introduction to the WIPP analysis, firstconsider a simple system of two layers sandwichedbetween a rigid top and bottom (see Figure A-I5). Thethicknesses, densities, and viscosities are h ll Pll /-L1l andh2, P2, /-L2 respectively, and the wavelength of thedeformation is oX.

The deflection of the interface is

y = y sin wx

where

w = 21f/oXoX = the wavelength

The rate of growth of the deflection is

v = V sin wx

and (following Ramberg) the associated horizontalrate of growth is

u = il cos wx

Shear and normal stress at the bottom interface of theupper layer are

+ pSg sin wx - p

Upper Layer

(el'?' ]Lower Layer

(~2,12J

Rigid Overburden11111/1111//11111111////1

t Interface Deformationhi y=ii Sin ~x

~

h2

I~ ).. ~I

7777~77777777777777777777Rigid Basement

y

Lx

Figure A-15. Simple Two-Layer Model

125

And at the top interface of the lower layer Again, dynamic equilibrium requires continuousnormal and shear stress at the interfaces betweenlayers. Material continuity requires continuous dis­placements and displacement rates. These boundaryconditions lead to a set of equations that, for a systemwith three interfaces, can be written

'.

where

p is the mean pressureS, P are constants dependent on the ratio of layer

thickness to deformation wavelengthg is acceleration due to gravity

Continuity of stress yields

where

This expression can be inverted to give

v = kqy

where k is termed the growth factor. The deformationthen grows exponentially with time

dy k ­dt = qy

y(t) = Yo ekqt

Plots of the growth factor versus wavelength typi­cally show a maximum at some wavelength. Thisparticular wavelength has the fastest rate of growthand, other things being equal, is the deformationwavelength most likely to develop in the system. Therate of growth is predicted by this maximum growthfactor.

The multilayer analysis is similar, but now theamplitude and rates of growth of all free interfacesmust be considered.

126

Cll C12 C13 C14 • • VI ...•

C21 C22 C23 C24 • • U1 q1YIC31 C32 C33 CM C35 C36 V2 •C41 C42 C4:j C« C45 C46 U2 q2Y2

• • C53 CM C55 C56 V3 •• C63 C64 C65 CGG u3 Q3Y3

where

Yi> Vi> Uj are the displacement and displacement ratesof the i'th interface

Cij , qi are constants dependent on the thicknesses,densities, and viscosities of the layers and onthe wavelength of the deformation.

This matrix expression can be expanded for systemswith more than three interfaces.

The preceding expression can be inverted and theterms rearranged to give

where

The elements of the k-matrix describe how a displace­ment on one interface affects the displacement rates,or velocities, on all interfaces. The total matrix repre­sents the coupling between the degrees of freedom(displacement rates) and loads (displacements) forthe system of layers. For such matrices it is useful toconsider the eigenvalues and their associated eigen­vectors.

The eigenvalues and eigenvectors are the con­stants K j i = 1 ... n and corresponding vectors fYh i =

1 '" n, which satisfy

I = 1 ... n

,..

where n is the number of interfaices in the system.Multiply by the constant ql

Recall that

ql [kl {y} Iv}

Then

where lVh is the time rate of change of the displace­ment eigenvector lYh.

The eigenvectors are then those particular dis­placements whose rates of growth throughout thesystem of layers are proportional to itself. These willbe the fastest deformation rates and are the ones mostlikely to develop. If the displacements are proportion­al to the eigenvector, then they will grow exponential­ly at a rate

WIPP ModelsThe basic WIPP-site density structure is fairly

well established (deviations from this basic structurewill be the subject of an interpretive report on thegravity survey). Figure A-16 is a copy of the densitylog of the Castile Formation in borehole AEC 8. Notethat Anhydrite II and Anhydrite III overlie less densehalite members. Figure A-17 is a simplified model ofthe overall density stratigraphy.

Density inversions occur in four places on thedensity model (Figure A-17). These are the Rustler/Salado contact, and the bases of Anhydrite III, Anhy­drite II, and Anhydrite I. Each of these interfaces isgravitationally unstable and would spontaneously de­form if the material were soft enough.

Seismic and borehole information indicates thatDZ deformations do not extend below the top of theDelaware Mountain Group. This is reasonably attrib­uted to the stronger, or less viscous, character ofclastic and carbonate sediments in the DelawareMountain Group. For the model analysis, the top ofthe Delaware Mountain Group is assumed to form arigid basement.

The Salado Formation was divided by an inactiveinterface at 2100 ft to determine the deformationeffects at the repository horizon. Densities and viscos­ities were constant across this interface.

The WIPP density model then has eight freeinterfaces. The associated k-matrix has eight eigenval­ue - eigenvector pairs, each of which can be related toa particular interface by inspection of the eigenvec­tors. The four pairs related to interfaces with stabledensity contrasts (density increases with depth) havepositive eigenvalues and, along with the negative q"describe an exponentially decaying or flattening de­formation. The one pair associated with the repositoryhorizon has a zero eigenvalue indicating no growth.This results from the lack of any density contrastacross this interface. The three pairs that relate tounstable density contrasts have negative eigenvaluesindicating deformations that grow exponentially withtime. These deformation modes will be examined ingreater detaiL

_ llA1AIlO _

fORMATIONANHYDRITE III ANHYDRITE II ANHYDRITE I BEll CANYON

FORMATION

F~g~re ~·16. Densilog of the Castile Formation in Borehole AEC 8 (The three massive anhydrite members are clearlydlstmgUlshed from the less dense halite.)

127

Olnait U (lm/cc)2.0 2.& 3.0I I I

•

Olwey Lake• Auetllr Formationa

1000 -

­•••--2000 -

3000

4000 -

Salado Formation

Anhydrite III

Ie

HeI... II 0

•CD

A ritl II E..0

I.L.

Halitl I •-.-••CDCo)

Antvtr1tl I

Oll.w.rl Mount.in Group

128

Figure A-17. WIPP Site Generalized Density Structure (These densities and thicknesses are used inthe two dynamic models.)

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

3.0685.5216.1415.5024.5173.6132.3391.5951.1450.85840.66560.53050.43240.35890.30270.25860.22340.1950

Anh IIIHalite I

2.8685.7298.395

10.3311.2411.249.7837.8776.2384.9724.0183.2962.7452.3161.9781.7071.4871.306

k-Matrix Eigenvalues(times -0.001)

Anh 1111Halite II

1.0242.0483.0724.0975.1206.1398.1139.872

11.2712.2112.7012.7912.5712.1311.5510.8910.209.510

Rustler!Salado

50100150200250300400500600700800900

100011001200130014001500

Deformation----;;".---...,....---,----:...-----;,...--,-~~--'--......,,........,,.....,--

Wavelength(m)

Table A-1. Growth Factors or k-MatrixEigenvalues for the Three GravitationallyUnstable Interfaces - Model 1

The equations relating normal and shear stressesto deformation rates depend on the wavelength of thedeformation. This dependency carries through theanalysis of a system of layers; the eigenvalues alsodepend on the wavelength of the deformation. Plots ofthe eigenvalues versus wavelength show a maximumfor each of the three unstable modes of deformation.These maxima are at the fastest growing deformationwavelengths; these wavelengths are thus the mostlikely to develop.

At the WIPP site, rock densities and layer thick­nesses are fairly well established by borehole and well­log data. Effective viscosities are much less controlledand are considered a variable in the analysis. Fortu­nately the analysis depends on the contrasts in rela­tive viscosity between layers. The absolute viscositiesare left out of the analysis until a final calculation ofthe exponential growth rate.

Two models were analyzed, both with identicallayer densities and thicknesses but dissimilar viscosi­ties. The first model assumes a constant (but unspeci­fied) viscosity throughout the entire section. The re­sulting plot of the three growth eigenvalues (for thethree unstable density interfaces) is shown in FigureA-18. The numerical values are given in Table A-I.

Figure A-18 indicates deformation wavelengths of150 m on the interface between Anhydrite II andHalite I, 300 m on the interface between Anhydrite IIIand Halite II, and 900 m on the interface between theRustler and Salado Formations. The correspondingeigenvectors are given in Table A-2.

0.014

0.012

~0.010

>

I 0.008

w>C

~ 0.006ce," 0.004

0.002

0.000 L-..L_-L_.L-.--=:l:::::r::=:I=:=::lt:::=..-1o 200 400 600 800 1000 1200 1400 1600

DEFORMATION WAVELENGTH 1m)

Figure A-18. Growth Factors or k-Matrix Eigenvalues forthe Three Gravitationally Unstable Interfaces - Modell

Table A-2. Normalized EigenvectorComponents at the Wavelengths ofMaximum Rate of Growth for the ThreeGravitationally Unstable Interfaces ­Model 1

Eigenvector Components

Rustler/ Anh 111/ Anh 11/Interface Salado Halite II Halite I

Free surface 0.024 * *Rustler/Salado 1.000 * *Repository 0.165 -0.005 *Salado/Anh III 0.006 0.140 0.001Anh III/Halite II -0.002 1.000 0.032Halite 11/Anh II * 0.152 0.159Anh II/Halite I * -0.011 1.000Halite I/Anh I * -0.005 0.020

129

1.2292.4583.6624.7285.5215.9886.1536.0795.8405.5025.1154.7154.3243.9543.6133.3023.0212.7692.5422.3392.1571.9941.8471.7141.5951.1450.85840.66560.53050.43240.35890.30270.25860.22340.1950

Anh II!Halite I

1.1472.2943.4414.5885.7296.8457.9028.8579.676

10.3310.8211.1411.3011.3311.2411.0610.8110.5010.159.7839.4019.0138.6268.2477.8776.2384.9724.0183.2962.7452.3161.9781.7071.4871.306

k-Matrix Eigenvalues(times -0.01)

Anh III!Halite II

Rustler!Salado

0.058520.11700.17560.23410.29260.35110.40970.46820.52670.58520.64370.70220.76050.81870.87660.93400.99071.0461.1011.1541.2061.2561.3031.3481.3911.5601.6531.6781.6521.5911.5101.4181.3241.2311.142

20406080

100120140160180200220240260280300320340360380400420440460480500600700800900

100011001200130014001500

Deformation----..~.,...---r-~:,.:...;.-..,..,.;.,....:..:...---.:--T__rT.,­

Wavelength(m)

Table A-3. Growth Factors or k-MatrixEigenvalue. for the Three GravitationallyUnstable Interfaces - Model 2

The first eigenvector in Table A-2 is for deforma­tions primarily related to the density contrast acrossthe Rustler/Salado interface. The deformation ampli­tude on the free surface should be about 2% of that onthe Rustler/ Salado interface. The deformation ampli­tude at the repository level should be about 16% ofthis interface, and the other interfaces should beessentially undeformed.

The second and third eigenvectors in Table A-2are for deformations related to density inversionswithin the Castile Formation. The deformation ampli­tudes on adjacent interfaces are about 15% of those onthe gravitationally unstable interfaces.

The eigenvectors show that the gravity-founder­ing mechanism should affect primarily the CastileFormation and the Rustler/Salado interface. The freesurface and intervening Salado Formation are expect­ed to show only minimal effects. From the plot ofeigenvalues versus deformation wavelength, the ex­pected wavelength at the Rustler/Salado contact isseveral times the wavelengths expected in the CastileFormation. Interestingly, the wavelengths are integermultiples of each other. This characteristic should beexpected in a dynamically coupled system.

A shortcoming of the constant-viscosity model isthat it predicts relatively large deformations on theRustler/Salado interface, while the observed deforma­tions are finite but small. The second model analyzedwas identical to the first except that the viscosity ofthe first layer (Dewey Lake and Rustler Formations)was ten times that of the underlying layers (Saladoand Castile Formations). This change seems justified;the first layer is primarily clastic sediments and thelower layers are evaporites.

For the second model, Figure A-19 is a plot of thethree growth eigenvalues (for the three unstable den­sity interfaces) versus the wavelength of the deforma­tion (Table A-3 contains the numerical values). Thecorresponding eigenvectors are given in Table A-4.These results are very similar to those of the firstmodel. The model predicts deformation wavelengthsof 140 m on the interface between Anhydrite II andHalite I, 280 m on the interface between Anhydrite IIIand Halite II, and 800 m on the Rustler/Salado inter­face. The eigenvectors predict minimal surface andmid-Salado (i.e., repository-level) deformation.

130

Table A·4. Normalized EigenvectorComponents at the Wavelengths ofMaximum Rate of Growth for the ThreeGravitationally Unstable Interfaces ­Model 2

Eigenvector Components

Rustler/ Anh 111/ Anh 11/Interface Salado Halite II Halite I

Free surface 0.027 * *Rustler/Salado 1.000 * *Repository 0.134 -0.004 *Salado/Anh III 0.001 0.128 0.001Anh III/Halite II 0.001 1.000 0.032Halite 11/Anh II * 0.142 0.148Anh II/Halite I * -0.012 1.000Halite 1/Anh I * -0.004 0.016

0.14

0.12

ANH. II!HAUTE II

!!l 0.10(K

mexat 280 m)

~>

i 0.08

>C0.08!,

II: 0.04

0.02

0.00 L...::::::::::L:"--L.._....L.-_.L:::::l:==:t:=:::::1==--.Jo 200 400 800 800 1000 1200 1400 1800

DEFORMATION WAVELENGTH (m)

Figure A·19. Growth Factors or k-Matrix Eigenvalues forthe Three Gravitationally Unstable Interfaces - Model 2

The primary difference between the results of thefirst and second model analysis is the relative rates ofgrowth on the Rustler/Salado interface. In the firstmodel the .maximum growth factor, or eigenvalue, atthe Rustier/Salado interface is larger than the maxi­mum growth factors of the two unstable interfaceswithin the Castile Formation. This model "predicts"large deformations on the Rustler/Salado interface. In.the second model, the maximum growth factor at theRustler/Salado interface is much less than the growthfactors of the two unstable interfaces within the Cas­tile Formations. This model "predicts" relatively small(and long-wavelength) deformations at this interface.Apparently the relative rates of growth can be adjust­ed by carefully selecting the relative viscosities.

Model results are shown diagrammatically in Fig­ure A-20. The wavelengths and relative amplitudesare those predicted by the second model, except thatthe deformation amplitude on the Rustler/Salado in­terface is intermediate between the two models. Theeffects of the three unstable density interfaces on allother interfaces are those given by the correspondingeigenvectors.

Rates of GrowthThe rate of growth can be expressed as the time

required for the deformation to grow e = 2.72 timesthe initial amplitude, or

where

-pzgh

2JLzpz 2.5 g/c3

, density of the upper layerg 980 cmM, acceleration due to gravityh 850 ft = 25 900 em, thickness of the upper layerMz = viscosity of the upper layer

KmaI is the eigenvalue, or growth factor, corre­sponding to the wavelength of maximum growth forthe particular interface considered. In the secondWIPP model the interface between Anhydrite III andHalite II has a maximum eigenvalue of 0.113.

131

1000 --- .----

.!!..<.>

c:.~

E:;...

Olwey Like• Ruetl.. Farmetiane

Se'eda Far met ian

Anhydrite III

Hllite I

Anlylrite I

Oellwl.. Mauntlin Group

Hlme II

_ Anhydrite II

1. i

I

1-

.~

IiII

-I

I

Rlpolitaly

ii!

-- - -I----~--I-- -.-'---- -.. --- ---- ..---

i-LI

-+---'---+--~-j---

2000 -+-----+---,--f-

4000 ­I

SurflCI - '

Figure A·20. Diagrammatic Presentation of the WIPP Dynamic Analysis (Deformation wavelengths and the effect onadjacent beds are predicted by the appropriate eigenvalue/eigenvector pairs. Amplitude on the Rustler/Salado interface isarbitrarily selected.)

From the preceding relations, the characteristictime is directly proportiortaltothe viscosity of theupper layer in the model. In the second WIPP modelthe viscosity of this upper layer (the Dewey Lake andRustler Formations) was assumed to be ten times thatof the underlying layers (Salado and Castile Forma­tions). Thus, assigning a value to the upper layerdetermines the viscosities throughout the entire mod­el.

The effective viscosity of large volumes of hetero­geneous evaporites under the actual conditions oftemperature, pressure, strain rate, and water contentthat exist during deformation is unknown. In relatingboth theoretical analysis and centrifuge models to

geologically reasonable growth rates of salt diapirs,Ramberg (p 97 and Chap 12) adopted 1018 P as a saltviscosity. Other authors have used different values.

Table A-5 shows the characteristic times for evap­orite viscosities of 1017

, 1018, 1019 P.

The rate of growth can also be approximated bythe velocity of deformation when the amplitude of thewaves is an arbitrary percentage of the wavelength(say 10%). For deformations on the interface betweenAnhydrite III and Halite II (Table A-5), the wave­length of maximum growth is 280 m. Waves with anamplitude equal to 10 % of this wavelength will growby an amount of 28 x (e-l) in one characteristic time.The resulting velocities are included in Table A-5.

132

Table A-5. Deformation Rates at theInterface Between Anhydrite III and Halite II(see text for details)

Gravitational FounderingThe analysis successfully accommodates a short­

wavelength deformation within the Castile Forma­tion; longer wavelength with less intense deformationat the Rustler/Salado interface; little or no deforma­tion at the free surface, at the repository horizon, andin the underlying Delaware Mountain Group. Themechanics of gravity foundering, rates of deformation,and evaporite viscosities are reasonably consistentwith the observations if a finite yield strength isassumed to retard deformation in the normal areas.

The analysis is limited to low-amplitude sinusoi­dal deformations in a system of infinite layers. Theboreholes, however, encounter vertical displacementsof hundreds of feet, strains of hundreds of percent,missing and duplicated units, lamina dips of flatthrough vertical (locally overturned?), solution/recrystallization and "possible faults. Clearly the DZ ismore complex than the case analyzed.

Gravity foundering will ultimately move the moredense material beneath the less dense. Centrifugemodels demonstrate the resulting structural complex­ity of this process. Such models are significant be­cause, if they are properly constructed, the model willpass through a structural evolution similar to thegeologic prototype. The suitability of the centrifugetechnique, the interplay between scaling ratios, andthe laboratory procedures are discussed in detail inRamberg's text (Chap 1 and 10 of the 1981 edition).

In a normal gravity field, gravitational tectonicmodels are difficult to construct because properlyscaled materials are too soft for convenient handling.

Centrifuge Models

•

Effective Viscosity (P)

Dewey Lake Salado andand Rustler CastileFormations Formations

1018 1017

1019 1018

1020 1019

Time(yr)

894389430

894301

Velocity(cm/yr)

0.5380.0540.005

A corollary is that large rock masses deform as verysoft or low-viscosity fluids over the extended durationof geologic time. The centrifuge technique replaces thenormal gravity field with very strong (2000 to 3000 g)centrifugal body forces. Various clays and putties thenpossess reasonably scaled properties for use in gravita­tional tectonic models.

The following models are taken from Chap 15 ofRamberg's text. They were originally intended todemonstrate gravitational deformations within anorogenic belt. However, if the scale ratios are rede­fined, the same models demonstrate gravity founder­ing of dense anhydrite through less dense halite.

The models were constructed from layers of paint­er's putty, silicone putty, and an oil-wax mixture incircular pans 10 cm in diameter. The density andviscosity of the layers differ, and the configurations ofthe layers vary between models. Figures A-21, A-22,and A-23 are radial sections of the models.

The models were mounted in a high-speed centri­fuge and subjected to the centrifugal loads noted inthe figure captions. They were then removed andsectioned. Figures A-24 and A-25 are photographs ofthe actual sectioned models. Figures A-26 and A-27are trace drawings of the deformations.

The models are all cases in which a density inver­sion occurs beneath a less dense overburden (the oil­wax layer). The density inversion is between a layer ofpainter's putty (p = 1.87 g/c3

) and the underlyingsilicone putty (p = 1.14, 1.25, 1.35 g/c3

). The thick­nesses of the dense layer are 3, 1.5, and 1 mm inmodels S112, S114, and S116, respectively. The densi­ties are given on Figures A-21, A-22, and A-23. Theviscosities are less well determined, but are given byRamberg (p 375) as 106

- 107 P for the painter's puttyand 3 x 105 P for the silicone putty.

Model scaling ratios follow from a comparison ofthe WIPP prototype with the models. If 1 mm in amodel corresponds to 100 ft in the prototype, then thesubsiding layers correspond to anhydrite members100, 150, and 300 ft thick. For a density ratio of 0.62,the corresponding anhydrite density is 3.02 g/c3 andthe corresponding "halite" layers are 1.84, 2.02, and2.18 g/c3• For a viscosity ratio of 10-12

, an averagemodel viscosity of 106 P corresponds to an evaporiteviscosity of 1018 P. Body forces and deformation timeare determined by the experiment to be about 2000 gand 10 min respectively. The duration of the proto­type deformation is determined by the interdepen­dent scaling factors to be around 700 000 yr. Thesescaling ratios are listed on Table A-6.

133

--------- 5em

/ I

~~3===~>1>- polnter',putty~ / 11.87g/eel

~. "'-gray ,lIIoon. 11.25 glee)

\ '--white aillcone(1.14 glee)

black ailicone (1.35 g/eel

5em -------_

;;

smcone putty (1.14 glee)\I

roil-wax (0.9 gleel

painter'. putty(1.87 glee)'

lilicone putty {1 14 g/eel

$ ;;oil-w.x 10.9 glee)

/

"-

-+~=I.ek ,moon. 11.35 gleel /

whit. ,lIIeen. (1.14 gleel

grey silicone (1.25 g/cc)

Figure A-21. Radial Section Through Centrifuge ModelS112. Run at 2000 g for 11 min.

Figure A-22. Radial Section Through Centrifuge ModelS114. Run at 1300-2200 g for 8 min.

r red silicone w/clav lave,. (1 14 glee)

/ r oil·wax 10.9 glecl

~~~~~~~~~~f~b§I.~e~k§'II~lcon~'~('~'3;5~g~/e~e¥):;~t:--painter', putty~ / 11.87 gleel\. "-- grey allloone (1.25 g/eel

"-- white llIIcone (1.14 g/ec)

--------- 5em-------->~

Figure A-23. Radial Section Through Centrifuge ModelS116. Run at 2400 g for 11 min, then 2900 g for 4 min.

Figure A-24. (From Ramberg, p 364)

134

Figure A-25. (From Ramberg, p 365)

Figure A-26. (From Ramberg, p 366)

135

Figure A-27. (From Ramberg, p 367)

Table A-6. Ratios for Reinterpreting Models S112, S114, and S116 asGravity Foundering of Anhydrite

136

Quantity

Thickness of thesubsiding layer

Body forces

Densities:Anhydrite

Halite(3 layers)

Viscosities

Time

Model

1 to 3 mm

centrifugalacceleration:

2000 g(approximate)

1.87 g/c3

1.14, 1.25,1.35 g/c3

106 P(approximate)

10 min

Prototype

100 to 300 ft

gravity

3.02 g/c3

1.84,2.02,2.18 g/c3

1018 P

7 X 105 yr

Ratio

i r = 3 X 10-5

ar = 2000

Pr = 0.62

t. = J.tr/Prirart. = 2.7 X 10-11

4.7 em

A quarter-of-a-million-barrel brine pocket is

tty

eling clay

\

\\\\\\ \\\ \' \

. ,,\, \'\ I '- \<

"'"~

¢~.. ~pa~r:u

u ua solution

Figure A·28. Radial Section Through Centrifuge Model M.Run at 210 g for 4.5 min.

Vp = 0.25 X 106 bbl = 4 x 1013 mm3

j

The mod.el consisted of a small volume of KMn04aqueous solution beneath interlayered modeling clayand painter's putty. The KMn04 stained the modelingclay and putty as the fluid migrated, providing arecord of the fluid path. Figure A-28 is a radial sectionthrough the model.

The model was subjected to centrifugal forces of210 g for 4.5 min. The fluid penetrated the surface in3.75 ±0.75 min. Figure A-29 shows a series of slicesthrough the model. The fluid appears to have followeda complex system of anastomosing channels.

Extrapolating this model to the migration of brinethrough the evaporites requires establishing modelratios for length, viscosity, density, and acceleration.The model ratio for time follows from these otherratios.

The length ratio is established by comparing thevolume of fluid in the model with the volume of abrine pocket. The brine pocket is here assumed tocontain a quarter of a million barrels. The model fluidvolume is

Some notes about the models:

1. The structural complexity of the models iscomparable with the more deformed portionsof the disturbed zone.

2. Some of the subsided units have remainedintact (e,g., model S116).

3. The deformation at t~e free surface is muchless than ill the underlying materials because ofthe stabilizing density stratification of the air/rock interface.

4. The less dense silicone clay layers are terminat­ed at:a vertical interface with more densepainter's putty. This boundary should influ­ence the style of deformation.

5. Scaling factors, the resulting evolution time,and deformation wavelengths are .reasonablyconsistent with the WIPP site analysis.

Migration of. Brine PocketsBrine pockets are irregular volumes of fluid en­

countered within the Castile Formation in areas ofcomplex structures. The fluid pressures are interme­diate between hydrostatic and lithostatic heads andare distinctly higher than the heads in adjacent aqui­fers (implying isolation from these aquifers). Theorigin of the brine is unknown.

Some of Ramberg's centrifuge models have a pos­sible bearing on the migration of brine pockets. Thismaterial is included here because the process is a formof gravitational deformation, and the interpretation issimilar to that in the preceding section.

The application of these models to the WIPP siteis equivocal. Particularly relevant is whether the brineplays a dynamic role in creating its reservoirs. If thereservoirs result from extension fractures in foldedanhydrite and the brine passively fills these fractures,then the following material should not apply. Howev­er, the brine pockets may form by concentrating previ­ously disseminated fluids in low stress regions withinthe developing structures. Then the reservoir volumewould be created or held open by the fluid pressurewithin the brine. In this case, the density differencebetween the brine and its host rock creates a tendencyfor the brine to rise through the section. Whether itdoes rise (and how fast) depends upon effective viscos­ities, yield strengths, and physical dimensions.

The model considered here is described in detailin Chap 13 of Ramberg's text (model M). It wasoriginally constructed to represent the rise of lessdense fluid magma through a heavier overburden but,by redefining the prototype dimensions, it appliesequally well to the ascent of brine through evaporites.

137

Figure A-29. (Ramberg, p 334)

138

The volume ratio is

and the length ratio is

i r = flv r = 5 X 10-4

The viscosity ratio is uncertain by several ordersof magnitude but can be estimated. The viscosity ofthe modeling clay is between 0.5 x 108 and 7.4 x 108 P.The painter's putty varies between 105 and 108 P withvalues up to 108 P when the strain is large and thestress is of the order of 106 dyne/cm2

• For thesecalculations a uniform model viscosity of 108 P isassumed. Salt viscosity has been experimentally de­termined to be around 1017 P, and 1018 P was found togive geologically reasonable rates of salt dome evolu­tion in models (0.2 cm/yr, Ramberg, Chap 12). Wetsalt may be much less viscous (Wenkert, 1979). Forthese calculations a uniform evaporite viscosity of1018 P is assumed. The viscosity ratio is then

/-Lr = /-LmodeI!J-Lprototype = 10-10

The fluid viscosity of the model (10- 2 P) thencorresponds to a prototype brine viscosity of 108 P.While this viscosity is much larger than the actualbrine, the contrast between it and the evaporitesshould be sufficient to believe that the model fluidcorresponds to a mechanically inviscid fluid in theprototype.

The density of the model fluid should be near1 g/c3 (it is not specified in the text). Brine density isnear 1.2 g/c3

; the density ratio is thus

Pr = PrnodeI!Pprototype = 0.83

The density of the modeling clay (1.67 to 1.71 g/c3)

then corresponds to prototype densities of 1.93 to2.05 g/c3 and the painter's putty (1.8 to 1.9 g/c3

) to2.16 to 2.28 g/c3

• This is reasonably close to the densityof rock salt.

The ratio of the body forces is just the centripetalacceleration expressed in "g," or (for this model)

ar = 210

•

The time ratio follows directly from the ratios oflength, viscosity, density, and body force as

Substituting

J.Lr = 10-10

Pr = 0.83£r = 5 X 10-4

ar = 210

then

t r = 1.1 X 10-9

The model time necessary for the solution topenetrate the overburden (3.75 min) corresponds to6500 yr. The 28 mm of model overburden correspondsto 56 m in the prototype; thus the prototype brinevelocity is around 1 em/yr. The time required topenetrate 3000 ft is then 105 yr.

DiscussionThe analysis and centrifuge models demonstrate

how DZ deformations can result from gravity founder­ing. These results follow directly from known thick­nesses and densities of the stratigraphic section andfrom reasonable estimates of the effective viscosity.DZ deformations are the expected result of the exist­ing physical situation.

The question is not why the DZ developed, butwhy the entire Delaware Basin has not similarly de­formed. A possible explanation lies in the dependencyof gravitational tectonic stress on the amplitude of the

density structure. In the preceding theoretical analy­sis of gravitational instabilities in a layered system, itwas noted that the velocity or deformation rate isproportional to the amplitude of the existing deforma­tion. Because the gravitational stress is linearly de­pendent on both the amplitude and deformation rate,small deformations imply small gravitational tectonicstresses. No deformation would occur if the evaporitespossess a finite yield strength that exceeds the smallgravitational tectonic stress associated with naturallow-amplitude sedimentary structures. Secondaryprocesses, such as external tectonic faulting, couldproduce an initial structure whose inherent gravita­tional tectonic stress exceeds the yield strength. Thedeformation would then grow outward from this area.An indication of such initiating structures is the struc­tural high at the top of Anhydrite I at boreholeWIPP 11.

An alternate interpretation for the localization ofcomplex structures within undeformed areas is sug­gested by the role of water on evaporite rheology.Wenkert (1979) noted that strain rate in Iranian saltglaciers was very much larger than predicted by previ­ously investigated halite flow mechanisms. He attrib­uted this to interstitial saltwater solution. D. Borns(pers. comm.) has identified low-temperature pres­sure solution facilitated by water as the primary de­formation mechanism in the DZ. Anomalous waterconditions are supported by the occurrence of brinepockets in disturbed structures in the northern Dela­ware Basin. In this interpretation, the complex struc­tures form in areas where anomalous intergranularwater facilitates grain boundary pressure solution.

If the DZ resulted from gravitational foundering,then the driving force will remain in effect until all theanhydrite has settled beneath the less dense halite. Inthis sense, the deformation should be regarded asactive. However, the deformation has minimal effectat the level of the proposed waste repository, and therates of deformation are slow.

139

Referencesc. A. Anderson, "Simplicity in Structural Geology," in

C. C. Albritton, Fabric of Geology (Geological Society ofAmerica, Reprinted 1963, pp 175-183).

R. Y. Anderson, "Correlation and Structural Relation­ships of the Castile-Salado Evaporite Sequence in ERDA#6, AEC #7, and AEC #8 Boreholes, Eddy and Lea Coun­ties, New Mexico," report to Sandia Laboratories, Septem­ber 1976.

R. Y. Anderson, "Deep Dissolution of Salt, NorthernDelaware Basin, New Mexico," report to Sandia Laborato­ries, January 1978 (rev May 1978).

D. Borns, Sandia National Laboratories, personal com­munication, January 8, 1982.

T. C. Chamberlin, "The Method of Multiple WorkingHypothesis," Science 15:92 (1890) (reprinted in 1965,148:754).

M. T. Halbouty, Salt Domes (Houston, TX: Gulf Pub­lishing Co., 1979).

D. W. Kirkland and R. Y. Anderson, "Microfolding inthe Castile and Todilto Evaporites, Texas and New Mexico,"GSA Bulletin 81:3259-3282 (1970).

H. Ramberg, Gravity, Deformation and the Earth'sCrust (2nd ed; New York, NY: Academic Press, 1981).

H. I. Snider, "Stratigraphy and Associated Tectonics ofthe Upper Permian Castile-Salado-Rustler Evaporite Com­plex, Delaware Basin, West Texas and Southeast New Mexi­co," (PhD diss, University of New Mexico, 1966).

D. D. Wenkert, "The Flow of Salt Glaciers," GeophysRes Ltrs 6(6):523-526 (1979).

140

•

DISTRIBUTION:

US Department of Energy, Headquarters (3)Office of Nuclear Waste ManagementWashington, DC 20545Attn: Larry Harmon, Program Manager (WIPP)

Wade Ballard, Director,Division of Waste Isolation (2)

US Department of Energy, HeadquartersOffice of Defense Waste & By-ProductsWashington, DC 20545Attn: Dr. Goetz Oertel, Director (NE-320)

US Department of Energy (3)Albuquerque OperationsPO Box 5400Albuquerque, NM 87115Attn: J. M. McGough, Mgr, WIPP Project Office (2)

D. Jackson, Director, Public Affairs Division

US Department of EnergyCarlsbad WIPP Project OfficeRoom 113, Federal BuildingCarlsbad, NM 88220

US Department of Energyc/o Battelle Office of Nuclear Waste Isolation505 King AveColumbus, OH 43201Attn: Jeff O. Neff

Battelle Memorial InstituteOffice of Nuclear Waste Isolation505 King AveColumbus, OH 43201Attn: S. Goldsmith, Mgr, ONWI Library

Battelle Memorial InstituteProject Management Division505 King AveColumbus, OH 43201Attn: Neal Carter, General Manager

Bechtel National, Inc. (2)Fifty Beale StPO Box 3965San Francisco, CA 94119Attn: D. L. Ledbetter

Dale Roberts

D'Appolonia Consulting Engineers (10)2340 Alamo, SEAlbuquerque, NM 87106Attn: Dev Shukla

Westinghouse Electric Corporation (3)PO Box 40039Albuquerque, NM 87196Attn: R. K. Brown

R. Jones (TSC)C&C File, c/o Ginger Wilkenson

Frank L. Parker, ChairmanDepartment of Environmental and

Water Resources EngineeringVanderbilt UniversityNashville, TN 37235

Konrad B. Krauskopf, Vice ChairmanDepartment of GeologyStanford UniversityStanford, CA 94305

Karl P. Cohen, Consultant928 N California AvePalo Alto, CA 94303

Neville G. W. CookDept of Material Sciences and EngineeringUniversity of California at BerkeleyHearst Mining Building, #320Berkeley, CA 94720

Fred M. Ernsberger250 Old Mill RoadPittsburgh, PA 15238

Richard R. ParizekDepartment of HydrogeologyPennsylvania State UniversityUniversity Park, PA 16802

D'Arcy A. Shock233 VirginiaPonca City, OK 74601

141

DISTRIBUTION (cont):

John W. WinchesterDepartment of OceanographyFlorida State UniversityTallahassee, FL 32306

John T. HollowaySenior Staff Officer2101 Constitution Ave, NWWashington, DC 20418

University of California at Los AngelesDepartment of Earth and Space SciencesLos Angeles, CA 90024Attn: J. Rosenfeld

The University of TexasDepartment of Geological SciencesPO Box 7909Austin, TX 78712Attn: J. K. Warren

California State University Long BeachDepartment of Geological Sciences1250 Bellflower BoulevardLong Beach, CA 90840Attn: J. G. Dennis

Dr. U. A. PfirterUniversitlit BaselGeologisch-paliiontologiches InstitutBernoullistrasse 32CH-4056 BaselSwitzerland

WIPP Public Reading RoomAtomic Museum, KAFB EastAlbuquerque, NM 87185Attn: Ms. Gwynn Schreiner

WIPP Public Reading RoomCarlsbad Municipal Library101 S. Hallagueno StCarlsbad, NM 88220Attn: Lee Hubbard, Head Librarian

Thomas Brannigan Library106 W Hadley StLas Cruces, NM 88001Attn: Don Dresp, Head Librarian

142

Roswell Public Library301 N Pennsylvania AveRoswell, NM 88201Attn: Ms. Nancy Langston

Hobbs Public Library509 N Ship StHobbs, NM 88248Attn: Ms. Marcia Lewis, Librarian

State of New Mexico (2)Environmental Evaluation Group320 Marcy StPO Box 968Santa Fe, NM 87503Attn: Robert H. Neill, Director

NM Department of Energy & Minerals (2)PO Box 2770Santa Fe, NM 87501Attn: Larry Kehoe, Secretary

Kasey LaPlante, Librarian

New Mexico State GeologistPO Box 2860Santa Fe, NM 87501Attn: Emery C. Arnold

New Mexico State LibraryPO Box 1629Santa Fe, NM 87503Attn: Ms. Ingrid Vollenhofer

New Mexico TechMartin Speer Memorial LibraryCampus StSocorro, NM 87801

Zimmerman LibraryUniversity of New MexicoAlbuquerque, NM 87131Attn: Ms. Zanier Vivian

USGS, Water Resources Division (2)505 Marquette, NWWestern Bank Bldg, #720Albuquerque, NM 87102Attn: J. W. Mercer

;;

..

DISTRIBUTION (cont):

USGS, Conservation DivisionPO Box 1857Roswell, NM 88201Attn: W. Melton

USGS, Special Projects Branch (2)Federal Center, Bldg 25Denver, CO 80225Attn: R. P. Snyder

NM Bureau of Mines and Mineral Resources (2)Socorro, NM 87801Attn: F. E. Kottlowski, Director

Klaus KuhnGesellschaft fuer Strahlen-undUmweltforschung MBH MuenchenInstitut fuer TieflagerungBerliner Strasse 23392 Clausthal-ZellerfeldFederal Republic of Germany

Klaus Eckart MaassHahn-Meitner-Institut fuer KernforschungGlienicker Strasse 1001000 Berlin 39Federal Republic of Germany

Michael LangerBundesanstalt fuer Geowissenschaften

und RohstoffePostfach 510 1533000 Hannover 51Federal Republic of Germany

Helmut RothemeyerPhysikalisch-Technische

BundesanstaltBundesalle 1003300 BraunschweigFederal Republic of Germany

Rolf-Peter RandlBundesmisterium fuer Forschung

und TechnologiePostfach 200 7065300 Bonn 2Federal Republic of Germany

Fenix & Scisson, Inc.3170 W Sahara AvenueSpanish Oaks D-12Las Vegas, NV 89102Attn: J. A. Cross

Gayle Pawloski, L-222Geologist CSDPLawrence Livermore LaboratoryLivermore, CA 94550

US Nuclear Regulatory Commission (3)Division of Waste ManagementMail Stop 69755Washington, DC 20555Attn: J. Martin

M. BellH. Miller

Dr. Gary L. DowneyCenter of Waste Management ProgramsDepartment of Social SciencesMichigan Technological UniversityHoughton, MI 49931

Dennis W. PowersDepartment of Geological SciencesUniversity of Texas at EI PasoEI Paso, TX 79968

7111 L. J. Barrows (6)7133 R. D. Statler7135 P. D. Seward9700 E. H. Beckner9730 W. D. Weart9731 A. R. Lappin (5)9731 D. J. Borns (15)9731 D. D. Gonzalez9731 S. J. Lambert9731 K. L. Robinson9731 S. E. Shaffer9731 C. L. Stein9732 Sandia WIPP Central Files (12)9753 J. C. Lorenz8214 M. A. Pound3141 L. J. Erickson (5)3151 W. L. Garner (3)3154-3 C. H. Dalin (25)

For DOE/TIC (Unlimited Release)

143