IIIIJ_IIIIINIIII1_ttiit'_

DEPARTMENT OF THE INTERIOR

u.S. GEOLOGICAL SURVEY

U.S. Geological Survey Committee for the Advancement of

Science in the Yucca Mountain Project Symposium on "Fractures_

Hydrology, and Yucca Mountain": Abstracts and Summary

by

Joan Gomberg 1, editor

Open-File Report 91-125

This report is preliminary and has not been reviewed for conformity with U.S. GeologicalSurvey editorial standards or with the North American Stratigraphic Code. Any use oftrade, product, or firm names is for descriptive purposes and does not imply endorsementby the U.S. Government.

1USGS, MS 966Box 25046Denver Federal CenterDenver, Colorado 80225

31

UIS'TSIHllliLl_l OF THIS i]OCUMENTIS {Jt_LI_IIrLU

U.S. Geological Survey Committee

for the Advancemelit of Science in

the Yucca Mountain Project

Symposium onI

"Fractures, Hydrology, and YuccaMountain"'.

Abstracts and Summary

September 13 & 14, 1990Denver, Colorado

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United StatesGovernment. 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 responsi-bility for the accuracy, completeness, or usefulness of any information, apparatus, product, orprocess disclosed, or represents that its use would not infringe privately owned rights. Refer-ence herein to any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof. The viewsand opinions of authors expressed herein do not necessarily state or reflect those of theUnited States Government or any agency thereof.

!

Table of Contents

' ' PageStatement of Objectives by Devin Galloway.; ............................................ 1

Summary. of Talks, Posters, Discussions, and Outstanding Issuesby Joan Gomberg ............................................................................ 2

Symposium Agenda ......................................................................... 6

List of Posters ................................................................................ 8

Abstracts:

Frc.c::. _.Sc_.ing of Frac._._re.Ver,vcrks in Rock by C_.,.,_stoplaerBarton .............. 9

Some General Properties of Joints and Joint Networks in HorizontallyLa_'ered Sedimentary and Volcanic Rocks -- An Overview by Earl Verbeekand Marilyn Grout .......................................................................... 10

Seismic Imaging for Fracture Characterization and Structural Definition byE.L. Majer, J'.E. Peterson, L.R. Myer, T.M. Daley, K. Kar'asaki, and T.V. McEvilly .................................................................................. 13

"Regional-scale Veloci_ Anisorropy Inferred from Nevada Test Site NuclearTest P.arrivals bv Stephen Harmsea ..................................................... 14

Fracture Counts from Borehole Logs by P.H. Nelson, R. Snyder, and I.E.Kibler ......................................................................................... 15

Fracture Studies in the Welded Grouse Canyon Tuff: Laser Drift of theG-Tunnel Underground Facili_, Rainier Mesa, Nevada Test Site, Nevadaby S.F. Diehl, M.P. Chornack, H.S. Swolfs, and I.K. Odum ....................... 16

Distribtwion of Fracture-lining Minerals at Yucca Mountain by B. Carlos,D. Bish, S. Chipera ......................................................................... 18

Stress Field at Yucca Mountain and in Surrounding Regions by loarmStock .......................................................................................... 20

Field Description and Measurement of Mesoscopic Faults with Appffcationto Kinematic and Paleostress Analyses by Scott M.inor ................................ 21

Rhyolite Flow Fields as Barriers to Paleohydrologic Flow in Eastern Nevadaand Western Utah Between Latitudes 37°30"N and 30°N by R. ErnestAnderson and Theodore Barnhard ........................................................ 23

Saturated-Zone Ground-Water Flow at Yucca Mountain, Nevada: Can FractureFlow Be Adequately Characterized? by I. Czamecki and A. Geldon ................ 25

Modeling of Flow and Transport in Fracture Networks by K. Karasaki,A. Davey, .1".Peterson, K. Hestir, and .1'.Long ......................................... 27

ii

Fluid Flow in Rough-Walled Fracntres by Robert Zimmerman, SunilKumar, and Gudmundur Bodvarsson .................................................... 28

,' b

Modeling of Flow and Solute Transport through a Variable Aperture PartiallySaturated Fracture by Richard Healy and Edward Kwicklis .......................... 29

Flow Through Variable Aperture Fractures by John Smoot .......................... 31

Void Structure and Flow in Single Fractures by Larry Myer ......................... 32

List of Attendees ............................................................................. 34

iii

Statement of Objectives

by

Devin Galloway

U.S. Geological Survey2800 Cottage Way, Rm W2239

Sacramento, CA 95825916.978.4648

Statement of S.vmposium Objectives

The principal objective of this symposium is to review the available

informm:ion on fractured/faulted terrains in terms of a coherent hydrogeologic model of

_ound-water fluid flow madtransport, particularly as it pertains to the Yucca Mountain

region. This review addresses the influence and si_ificance of fractures on _ound-

water flaw and the transport of conservative-species solutes within the context of the

hydrogeolo_c setting of the Yucca Mountain area. The relations between fluid flow

.... and fractured or faulted host rock are examined integrally from information on

geologic, seismolo#c, hydrologic, and geomechanicaI properties of the system. The

development of new hydrogeologic approaches that incorporate information from this

int%m_.teddatabase arc contrasted with more standard approaches toward understandingflow in fractured reservoirs.

Ground-water flow in both the unsaturated zone and the saturated zone are

considered. The application of various models of flow is addressed; examples include

porous-media equivalent and discontinuum fracture-network models. Data and

interpretations from the Yucca Mountain area are presented to establish a context for

information exchange. The symposium includes discussions relevant to technical

considerations for characterizing the Yucca Mountain area hydmgeology. On the basis

of these discussions, CASY has compiled this document in order to formally

summarize the proceedings and communicate recommendations for future directions of

research and investigation.

gev.rSEb P_6_ . _-z_o_5

Summary of Talks, Posters, Discussions, Outstanding issues

byJoan Gomberg

Branch of Geologic Risk AssessmentUS Geological SurveyBox 25046, MS 966

Denver Federal CenterDenver, Colorado 80225

303-236-4410

This introduction summarizes the information presented at the symposium and some

of the outstanding gaps in our understanding of the relationship between fractures and the

hydrologic system. Fractures refer to adjacent surfaces in a rock where brittle failure has

occurred. When there is only shear motion on a fracture it may be called a fault, and often

the term fault also implies that the fracture has length of the order of a km or more. Joints

typically refer to fractures with no shear and possibly normal motion across the failure

surf'ace. The symposium was appropriately opened with the challenging proposal that the

nature of the problem at hand is such that it will never be possible to predict details of

future hydrologic systems (Neuzil talk). Rather, it may only be possible to define an

envelope of possibilities. If this is the case, the shape and size of such an envelope is as

yet unknown so that further studies are required to determine if this view is in fact

justified and whether such an envelope is sufficiently small to be useful. Alternatively,

a somewhat less bleak view is that we are acquiring knowledge at a rate too slow to be

able to adequately address the hydrologic issues facing us at Yucca Mountain but that thereis no fundamental limitation to what we can learn. If this is the case, or if one views the

potential for scientific progress more positively, then continued investigations must be

pursued. In other words, whatever one's views of the state of the science are, further

study is essential.

In contrast to the lack of consensus on the state of the science, there was near

unanimity among participants that progress to date has been far too slow. This sentiment

was expressed by symposium participants in the wrap-up session and in informal

discussions. This slow progress is evidenced by the lack of recent work presented by the

Yucca Mountain Project (YMP) participants and the dearth of new data from the vicinity

of Yucca Mountain. The lively scientific debate inspired throughout the symposium

strongly suggests that significant scientific progress will be made only when data collectionin the field is restored.

Fracture descriptions have yielded some important insights into their generation and

evolution. The application or fractal theory to fractures (Barton abstract) demonstrates that

fractures follow fractal scaling laws over many orders of magnitude in length; this suggests

that a single physical process may govern all fracture phenomena. Another implication of

fracture generation being governed by a fractal process is that one fracture influences

another. Thus, it is important to study a set of fractures collectively and to conduct studies

of fractures with scales ranging from microfractures to faults. A statistical study of joint

characteristics (Verbeek and Grout abstract) also has demonstrated that a systematic

approach to the description of fractures can yield an understanding of their generation

(e.g. lithologic control, effects of temporal and spatial sequence in a suite of fractures).

Although some work is being done (Minor abstract), much lesser emphasis has been placed

on systematically and quantitatively describing characteristics of faults.

It is now possible to make direct observations of fractures at shallow depths (less than

a few kilometers). Borehole logs from Yucca Mountain indicate that the relationship

between fracture count (derived from television logs) and flow (as inferred from resistivity

logs) is not simple (Nelson et al. abstract). Hydraulic tests in boreholes coupled with

fracture data from televiewer logs from the same holes also indicate a complex relationship

between degree Of welding, fracture distribution and interconnectedness and ground water

flow direction, magnitude, and travel time (Czarnecki and Geldon abstract). In addition

to serving as windows to view fractures in situ, the use of boreholes provides core samples

to examine the fine details of fractures ( Carlos et al. and Diehl et al. abstracts).

Mineralization in microfractures can serve as keys to paleo flow patterns. Microfractures

observed in core samples from the unsaturated zone (Grouse Canyon tuff) are often lined

and sealed shut with hydrous mineral phases (Diehl et al. abstract). In situ stress

measurements also have been made in the holes left from the core sampling providing

estimates of the present state of stress. The stress variations can be interpreted as arising

from lithologic differences (e.g. welded vs non-welded tufts) and are consistent with the

abundance and orientation of fractures. Thus, one might conclude that the paleo stress

conditions were similar to the present stress conditions, that they are/were conducive to

the opening of microfractures, and that the hydrologic conditions were quite different in

the past such that water flowed trough the fractures resulting in significant mineralizationwithin the microfractures.

Detailed analyses of the distribution of various minerals lining microfractures observed

in cores from and very near Yucca Mountain provide evidence of a horizontal

stratigraphic control on the paleo flow system, The mineral types and their sequencing

serve as clues to the evolution of past water chemistry and thus, possible paleo flow paths

and hydrologic conditions. Many of the minerals appear unaltered, suggesting that little

water has flowed through the fractures since the last episode of mineralization occurred.

However, the observations also suggest that microfractures have undergone reactivation

and therefore it is possible that they might operate as viable future transport paths (Carlos

abstract) should the climatic conditions change. Attempts to determine the age of

mineralization must be made to fully define paleo flow patterns, and further petrographic

analyses must be conducted to better define the lateral extent and abundance of

mineralization. Apparently many valuable core samples exist but are not available for

analysis, thereby making it nearly impossible to properly address these key issues.

Less direct observations of fractures are now being made using seismic methods

(vertical seismic profiling, cross-hole tomography, local earthquake/explosion travel time

delay analyses). These are proving to be viable tools for mapping fracture distributions on

useful vertical scales (e.g., of a single stratigraphic unit such as Paintbrush Tuff at NTS;

Majer et al. abstract) and horizontal scales (e.g., to resolve variations between Yucca

Mountain and Crater Flat; Harmsen abstract). However, because the methods are indirect

and thus interpretation of results is inherenltly non-unique application of such methods

must be combined with other independent types of studies.

The generation of fractures is dependent on the local and regional stress field and on

the pre-conditioning of the rocks (e.g., existing fractures, rheologic properties). In situ

stress measurements from Yucca Mountain show that stresses are close to those necessary

for failure of optimally oriented normal faults and that there is a dynamic relationship

between the stress conditions and the water table level (Stock abstract). The opening of

fractures may serve as new conduits for flow causing a rise in the water table. Conversely,

a rise in the water table due to climatic changes may result in opening of fractures thereby

resulting in a tectonically induced change in the water table. Fault control on the paleo-

hydrologic system is evidenced in the Eccles basin in eastern Nevada and in the Hamlin

Valley of western Utah (Anderson and Barnhard abstract). In both areas faults appear to

have acted as the major pipelines to the surface for transport of carbonate-rich water from

in a deeply circulating regional ground water system. The strong relationship between

I

watertable,stressconditions,and/orfaultingisalsomadeevidentbythechangesobserved

inthehydrologicsystemfollowingtheLoma Prietaem_hquake(Rojstaczertalk).However,

the physicsgoverningthisinteractionare not understood(Rudnickitalk)and the

distributionandcharacteristicsofseismogenicfaultsarenotwellknown,

The tremendousimportanceof fracturesas controls on hydrologic flow s_msto bean accepted scientific premise (Czarnecki and Geldon abstract), The advent of fast

computers makes it possible to construct complex models of flow through a single fracture

(Healy and Kwicklis, Smoot, and Myer, abstracts) or through a more complex system of

fractures (Zimmerman et al., Karasaki ¢t al, abstracts; Kelkar, C_ecki talks), in order

for these complex models to be useful as predictive tools, there must be an adequate

amount of data to constrain them. The major source of uncertainty in modeling studies

now seems to arise from a lack of adequate data rather than a poor understanding of the

fundamental physics and/or numerical techniques (Karasaki et al. abstract). It is now an

appropriate time to make serious efforts to coordinate modeling studies with data gathering

activities. We now have models of flow in the Yucca Mountain region (Czarnecki talk).

If these models are to be made useful, we must focus our efforts on timely gathering and

analyses of the relevent data and on the free exchange and innovative use of scientificinformation.

Symposium Agenda

September t3, 1990

7:_5-8:[5 Registration8:15-8:30 Announcements

8:30-9:00 I, Overview."The role of fractures in groundwater hydrology*', C. Neuzil (USGS-Fd

II. Fracture Generation & BehaviorA. Descrip_iy¢tools

9:00.9:30 i.Fractah-"Fractalanalysisoffracturenetworks",C.Barton(USGS-D)

9:30-I0:00 2.Statistics-"Statistical properties of r_a/fracture networks", E. Verbeek CJSGS-D)

I0:00-I0:15 Coffee Break

B. Obse_ationaI constraintsI. Field-

10:15-10:45 "Field description and measurement of mesoscopic faults

with special application to kinematic and pa.leostressanalyses', S. Minor (USGS-D)

IQ:45-Ii'15 2.Fracturemineralization-"Distnbudon of fractuze-lining minerals at Yucca Mm", B, Carlos (LANL)

l 1:15- I I' 4.5 3. Seismic anisotropy-"Seismic imaging for fracture chancterizadon andcrustal definition ', E. Majer tLBL)

11:45-1:00 Lunch

C. Models1:30-2:00 "Void structure and flow in single fractures", _ Myer (LBL)

.":00-2' 15 Coffee Break

III. Stress & Strain as Driving Functions for FlowA. Re_onal S_-ess/s_infi¢td

2:15-2:45 "The stress field at Yucca Mountain :andhasurrounding regions", J. Stock-/Harvard)

2:45-6:30 YMP PostersRefreshmentsInformal Discussion

September 14o 1990_:15-8:30 Recapof PreviousDay, R. WheelertUSGS-D)

_:30-9:00 "Hydrolo_cinformationsqueet.-'dfromtheresponseof porefluidstocrvst_deformattoi'L, S,Rojst=czer(Duke)

B. E_h__9:00-9:30 "E_nqu_es =adfluid flow ', 1,Rudnicici(_'orthwes|em_

IV. Flow ModelsA.M_.dt,.l_

9:30-t0:00 ",'vIodeblnilflow andsolute='_sportmmullhav_lble;Lpermre,p='ti_ly s_mratedt_ctu.fs', R,Healy'(USGS.D)

10:00-10:15CoffeeBreak

10:i5.10:45 "ModeLingofflowinfracturencr,vorks",K.K_M_ (LBL)I0:45-I I:I._ "Computer modelingof 3-Dcoupled heac.n_ss.s_'esseffe¢:s ', S,

K¢Lk.ax(LA,_"L.)11:15-I 1:J,5 "Flow tlvougttv_able aperture_=r=c=ures",1,SrnootiP_)

,, _5-1:00 Lunch

B.,x.o_|ic:l_ionstoYucc_Mount_in"1:30-2:00 ",Modeliagsainted'zone Foun(.i.wacerflow atYuccaMoun_

and vicinity',can we =dequ;_ceiysimui_¢efree.re flow?", },Cz:L,'neckJ(L'SGS-D)

"_:00-2:30 ApnhccL_onot discrece_:lcrue=o/_sts¢orepositorycha.,-actet',zatzon_d pertormaace=sessmen_,Tom _oe

(GolderAssoc,)=:30-2:45 Coffee Break

2:_5_:00 Wrap-Up. K. K_ras_kd(LBL'_,D, G_loway (USGS-S), D,GiiUes(USGS.D)

Abbreviations: USGS-D= USGSDenver, USGS-M = USGSMenloP_rk.USGS-R = USGS Reston. P._VL= Pac_J'ic,'_V Lab,LANL- Los Al_.raosNational L_b,LBL. L;zwrenc=Betklcy L_b, USGS-S =,USGS Sacr;m_en¢o

...... ii • II -- i

Posters i

"OSl$_ F__s atYuccaMountain"_wney, KQim,T_r, andErvtn.

__iw _ _ u _xlmmto paJ_ydroioricflow in e:LsteniNevadaandwesternUt_ It latitudeY,o30'N"E, Andenon,T, B_hud

_*a._r m_on m _.N_ ,_ffs,Yucc; Mountain,Nevaci_:Therecordfromursntumit_cUes"

R. Z_eUnslci

tJ,..qum_eonm_ velo_Ms,P.0J_ MesatoYucc:LMount._in" _

S.H_sen

_temUaml fine',urnt_tem leome_ fromwellcem"T_,Doe

_ _,'l_i_ O(I_rlo_¢ s_n wives:am2osphe_c,E_a'chtide,andseisrr;c"D, G_oway

_ sr_es inmeWeldedGrouseCanyonTuff:Luer dz'Lt'tof theG-TunneL__r_ f:,:_iry,P_nmr ,_{e.. _'eva_ TestSite,Hev_da"

_ $, D|ehl,51. Chom_ck,H. Swoifs.,r.Odum

'Fr_r_re countsfromboreho[e[oils'P, Nelson,R. Snyder,J, FAbler

"S_t,,n o( theupper_t (0.4krn)at Yucca_ount_"W, Mooney

___,Kum_, Bodvusson

_l,_p_.,,_me_ o(_ T_ma._withseis_c reflectionpro_es in C_er FIacandY_¢c._Moun_n'T. Brother

Fr'actalScalingof FractureNetworksinRock

Christopher C. Barton (U.5. Geological Su.rvey, Box 2..50._, MS 91.3, F'ederai Center, Denver, CO 80_225

The mathemadcai construct or"fractal geometry is well suited to quaad/y and model spadai and size-sc.a_g

reladom witkia complex systems that are stadsdcally self-similar over a broad range of scales. My results show

that natural fracture networks ha rock follow a f.ractal scaling law for fractures ranging in length over ten orders

of magnitude, from microfracrures i, tcctomicaily deformed quartz sad plagiodase grains to transform faults in

the South Atlmtic sea floor. Detailed measurements of two..dimeasionsI samples of three-dimensional fracture

nee'works in rocks of dissimiaar age, Iithology, md tectonic seeings show fractai dimensions in the range 1.6-1.3.

The small range is fi'actal dimension implies that a single ph.,_ical process of rock fracmrisg operates over rhis

wide range of scale, from microscopic cracks to large, iota'a-plate fault systems.

F_eld evidence has established that rock fracturing is as iterarive process in wimichpreexisting fracture.s influence

the format.ion of subsequent fractures (such behavior is chazacterisfic of fi'actal processes). Fracture networ_

are not random but evolve fi'om initially ordered to increasingly _'_sordered parterre, and they become more

_complex with tLme as new fracnlre generatiom axe added to those that already exist. The spadaI disr..,'ibution of

_rac_es w_fl_ the=e_'orkevolvesa_tracn_resaresequend_-!Iyadded to r_enet_vork.The fra_I dimcmsion,

which is a quantitative measureof thespadal distribution of thefracn_c trace.s, rang_ from about 1.33 hatearly

stages or"network development to 1._ [or mature networks. The fracr.al dimension of each successive fracture

generation is less than that of the pre=diog gcneradom but the fractal dimension for the cumulative network

in=eases as each new generation is added. The fractal behavior implies that fracture-network development is

governed by a nonlinear equation. Fortunately, the ability, of fi'actal mathematics to accurately quantify and

model the spatiaJasd sizc-sc.aling properties o( the .system is not dependent on specific knowledge of this

equadom

SO._{E_ PROPSRTIES OF JO_ AND JOLNT NE_,_OP_KSINHORIZONTALLY lAYERED S_]LMSNTARY A_NDVOLCANIC ROC_--

AN (Pa_R_flE_

Earl R. Verbeek and'_farilvnA. Grout

The past fifteen years have seen an unprecedented surge in the stud>'ofjoints, spurred largely by the need to understand the fluid-flow properties ofnatural fracture systems both to ensure :he safe storage of toxic andradioactive wastes and to more effectively recover gas and oil from fracturedreservoirs. Joints no longer are the eni=_matic features they once were; it isnow realized that their properties are related to lithology and to stratalsequence in consistent and understandable ways. The following discussionsummarizes some general properties of joints and joint ne_¢orks and is basedon recent work in diverse geologic settings of horizontal to gently tiltedsedimentary,and volcanic rocks.

Orientations

Strike dispersions of 15°-30° within individual sets are common; barringcomplications, the data often approxin_=e normal distributions. Smallerdis_ersions_ as little as 5o , have been documented for some early-formed setswithin fine-=trained,well-cemented, brittle rocks. Strike dispersions tend toincrease with increasing thickness of the jointed laver, with increasing grainsize, and especially with decreasing degree of induration. The greatestdispersions result from jointing of !ithologically heterogeneous units (e.g.,a variably cemented channel sandstone) over protracted spans of time duringwhich regional stresses progressively change in orientation; dispersions o=60°-70° have been doct=aen:edbut are uncommon. In such cases, too, strike-

frequency distributions for individual sets can be decidedly non-normal.0,_.._,any given iointed laver, strike dis.torsionsof the earliest set tend tobe least and of succeeding sets progressively greater.

Dips of joints within thin, planar-bedded, well-cemented rocks commonlyare within a few degrees of vertical. As with strikes, dip dispersions tendto increase with increasing thickness and grain size of the jointed laver andwith decreasin=_ degree of induration; dip dispersions of 20o-30°, and iocallymore, within thick (>_m), weakly cemented sandstones are not uncommon. Theinfluence of bedding on joint dip is seen in the many places where the dipdispersion for a given set is less than the corresponding strike dispersion.

Dimensions

Heights of joints are influenced primarily by the thickness of theindividual deposi_ional units in a stratigraphic sequence and by the degree oflithologic contrast be_¢een them. Small joints are characteristic of thinlybedded sequences of contrasting rock type (e.=_.,limestone or cher_ layersalternating with shales), whereas the opposite is true for thick, internallyhomogeneous units (e.g., massive sandstones and the massive upper parts ofsome ash-flow sheets) and within sequences characterized by low lithologiccontrast <e.g., some lacustrine siltstones). Far more common, however, arejointed s:ratal sequences characterized by variable bed thickness and moderatelithologic contrast between beds; common examples include stacked sequences ofpoint-bar sandstones and siltstones. Joints in such deposits show apredictably wide range in height, from small joints confined to individualsiltstone partings to large joints that cut several beds in succession; rangesof v.¢oto three orders of ,_a_nitudeare common.

10

. ,,eightand length is often observed,A rough corresDondenc_ between joint '-though few exposures offer sufficient three-dimensional control to quantifythe relation. Join= lengths for early-formed joints likely are related =omagnitude of driving stress, but for later sets the influence of pre-existingjoints becomes increasingly important. _Joint lengths for successive sets willtend to be progressively shorter if older joints remain open so that youngerjoints terminate against them, but no such relation need exist where olderjoints are healed and younger joints cut across them unimpeded. Here again,as in unfractured rock, driving stress ,-myplay a major role in influencingjoint length. Examples are known where younger joints are longer on averagethan those formed earlier.

Frequency distributions of join= length for mesoscopic joints oftenfollow a power-law distribution; i.e., progressively shorter joints arepresent in increasingly greater abundance than longer ones. The claim of acontinuum of the power-law relation to microscopic joints has not, to ourknowledge, been demonstrated for any area; mos= workers, f_r practicalreasons, have imposed an arbi:rary lower bound on the len=_thsof the fracturesthey measured and thus have left undocumented the mos_ cri'_ica!part of thedistribution. Field observations in several areas have sho_vntzat small

joints, those with trace lengths of several centimeters or less, are presentin lesser numbers than are larger joints of the same set, suggesting =ha_ theactual frequency distribution of joint lengths may not be a power-law functionfor short joints.

SpacingsSpacings of joints are influenced primarily by the litholog7 and

thickness of the jointed layer. For any given lithoiogy a s_ron=_relationber._eenmean (or median) joint spacing and layer thickness au the outcropscale often is evident. The functional form of the relation has been debated,but r°vi=w of older '--_" _ suggests that ecua=ions.,u._ su_-_!ementedkv .._uchnew wor_"

of the form log S = m log T * c, where S = _edian joint s_acins, T = layerthickness, and m and c are constants dependent on litholo_, hold for theentire range of S and T so far examined. E vs. T values for different jointsets in the same e.rposureplot as parallel but generally noncoincident linest>mt express quantitative!v the greater abundance of one set relative toanother for any given laver thickness. For different rock t]_es the S vs. Tvalues generally plot as lines of differen: slope; hence, the abundance ofjoints in one rock type relative to their abundance in another is itself afunc'.ionof layer thickness.

The above re!a=ions hold only where the boundaries between jointed layersare well defined and the individual joints span the full thickness of thelayer. IChere instead the lithologies are _rada=ional from one layer toanother, as among layers defined by different de_=reesand combinations ofwelding and devitrifica=ion within an ash-flow sheet, the join=s, like therocks, show gradually changing propezties ver=ically. Mechanical contrastswithin ash-flow sheets tend to be most pronounced in the lower portions of thesheet, where large differences in degrees of welding and devitrification occurover shot= vertical distances, and tend to be more obscure within the thick,

massive, par=ially,welded to nonwelded tuff that forms the upper parts of somesheets. Complications also arise within layers whose thicknesses are muchgreater than the heights of individual fractures. The joints in some suchrocks, particularly massive sandstones, tend to congregate within zones. Eachzone contains multiple, overlapping, closely spaced, and commonly in=croon"nec-.ed joints separated by broader intervals of less-fractured rock. Zonal

11

development of joints has been little studied and is incompletely understood.

Interconnections

Joints can terminate either by dying out within the rock, commonly astaoerin=_cracks decreasin_ gradually to z_ro aperture, or by abutting pre-existing fractures. In addition, younger fractures can intersect olderones. The relative proportions of "blind" endings, terminations, andintersections commonly are variable from set to set and are a function of theage of a set relative to others in the same rock, the mineralization historyof pre-existing sets, and to some de_ree the spacings of fractures alreadypresent. For obvious reasons blind endings are common and intersections rareamong members of the earliest join= set, but abutting relations can also beformed in abundance as later-formed members of a set "hook" into earlier onesto form so-called J terminations. Hence, fluid flow through interconnectedfractures can be effective in a rock even if only one set of joints ispresent. Joints of later sets will tend to abut ezrlier ones if these areopen but intersec',them if they have been "healed" effectively bymineralization. Variable proportions of abutting and crosscutting relationsare common where early joints are intone,lete!y mineralized and lenticular rugsremain, or :,herethe mechanical contrast bet'.veenmineral fill and wall rock is

sufficient to stop the propagation of some joints but not of others. In anycase, the degree of interconnection between younger and older joints tends tobe very hio_hunless the older joints are unusually sparse and thus widelyspaced within the rock. Poorly to moderately interconnected join: networks,in our e.rperience,are Fare.

Surface shapeEarly joints in many areas of relatively thick and homogeneous rock have

roughly elliptical surfaces with their lon=_axes parallel to bedding. Aslayer thickness decreases and progressive!'/more joints span the full_;-:'-° of the _=':=_ t_o joints pro=cressive!',azDroach a more rec'.anTa!arz....._..ss

form, with flat top and bottom edges. Join= surfaces in thinly beddedsequences of high lithologic contrast tend to be long and ribbonlike, quiteunlike their counterparts in thicker layers.

Joint surfaces of later sets exhibit a wide range in form dependent onlaver thicllness,abundance of fractures already present, and the dezree towhich these earlier fractures interfered with the propagation of the newerfractures through the rock. In relatively thick layers already cJt byabundant joints it is common for the ver'.icaldimensions of later joints toexceed the horizontal.

Cross-sectional shapeFrom field observations it is apparent that many early-formed joints in

relatively homogeneous rock have a maximum wall separation near theirmidpoints and gradually taper to zero width a= el=her end. Cross-sectionalshame as a function of joint size has been little studied, but for one area wehave shown t.hat_iI separations at join= midpoints are linearly related tojoin= heights (in a vertical cut) over the total height range of 1-360 cm.Join: shape in this area, then, is relatively constant regardless of jointsize, and the largest;joints possess by far the largest apertures.

No such relation seems to exist for joints of later sets. Instead, where

late joints terminate at both ends against open, pre-existing frac'.ures[freesurfaces), the _all separations of the later frac:ures may b_ roughly cons=an=along their length, in sharp contrast to ear!y-formed joints in the same rock.

]2

Seismic Imaging for Fracture Characterizationand Structural Definition

'

E. L. Majer, ,1. E. Pe:erson, L. R Myer, T. M. DaleyK. Karasaki, and T. V. McEviilv

Center.for Comptttational Seismolog3', Lawrence Berkeley Laboratory.,Ber/_.eiey, CalifOrnia 94720

V'SP and crosshoIe tomographic methods are being developed andtested as part of DOE's nuclear waste program for characterizing theYucca Mountain site. Work has been progressing in the develpoment ofmodels for understanding seismic wave propagation in fractured 3-dheterogeneous media and in Iieid testing fracture characterization methods.These field experiments have been utilizing high frequency (1000 to 10000Hz.) signals in a cross-hole configuration at scales of several tens ofmeters. Three component sources and receivers are used to map fracturedensity,, and orientation. The goal of the experiments has been to relate theseismo:logical parameters to the hydrological parameters, if possible, inorder to provide a more accurate description of a starting model for hydro-logical characterization. ResuLts of """.. ,.,.se controlled experiments indicatethat the fractures have a si,.,niricant..,effect on the propagation, of the P andS-waves. Laboratory experiments indicate that saturation also has adr',_,z,..aziceffect as '.veil. Work in,,'o[vinz d:=..verification of the sdffr,essw

theory indicates that t.he theory is va.iid and at high frequencies (greaterthan a few kilohe:'tz) the greatest effec= is on the amplitude, and at lowerfrequencies the greatest effect is on the velocity or delay of the seismicwaves. Lu addition to these controlled experiments, multicomponent VSPwork has been carried out at several sites to deten'nine fracture characteris-tics. The results to date indicate that both P-wave and S-wave can be

used to map the location of fractures, ha addition, fracture that are openand conductive are much more visible to seismic waves than non-conductive fractures. Recent work at the Nevada Test Site indicates thatthe Paintbrush Tuff is heterogeneous with respect to both P- and S-waveproperties. There is observed anisotropy in d'Leshear wave VSP data, butthe anisotropy s_,ems to be associated with near surface sediments (I00meters and less) and not associated with the Tuff.

13

?,._,_.:on-._.-c=,_','e:oc::y :,.'_:-o::'o._y.nier.',._ :':_ "_.v._d,,. -'._._c._i'.e :.ucie,..: ".es:P-_.r."lv_Li

The sou_"....G.-estB_._in-',",mi"net'.vor!<I_G "_'''............... ,, s:=::c,ns : ..",o_t['.,"','er:ic=i ¢omoonenc ,or _:vo compo-

nent t, "vuich _p_:_ an ,_pproxima:e]7 c:-cui_" :'eg:on o:"::,dius 160 :<rn. centered on ':'uc:_, ._[ounc_,in. shows_:'on__:imuch,% ',:r_:_ono_"?.w_ve :r:vei time dei._ys for Hev_d_ Test 5[:e {._'TS) ,'tucie_r device casts:t P:huce .,Lee:t, R_.i_Ler ._[es_. -nd Yucca. F!:c, Her_, delay is defined _ observed :ravel time minus

_heoret_c:.L _irne, computed :'tom a horizontally isotropic velocity model. For all Yucc: F!_t. R_inier _[es_,• nd .-_re_ 19 Pahu:e _,[es_ :es_s, :he t'_cest dL'ec:i'ons, correspond/n_ :o :he minimum de[._ys, ._re _bouc

north 10° - !O° e_: anti south _.0° - _9° west..he s/owes: dixecttons (m_.xu'num delays) ._ppe_x co be

_pproxirna:eiyper?endicu[_rto:hisorientation.One-second v_ri_tionso_"tr_.vei-cimedelaywith _imuch

_re observed for R=inier),[es._sources,forpaths h_vin_;lengthsless:h_n !00 krn. Y!_ure ].shows :he

_:zmu:h_L,iis:rtbutionoi _GES:; _ta:ion dei:tysfor :ire _:,inierL[esa. _est "D[sko Elm _ _ open cLrc_.es(_llRainier),les._nuclearrestdeiaypatterns._reverysim_i_-).The iunc:ionvalues,_ c_s:(_- l0°),b, where

._= sourceto sc_tion::irnu:h,.,hown_s soiid:ri_n_jiesforcornp._risonwith observed del0,ys(correlation

toe,cleric,_ =_0._),are, to _rs:order,_heore_icaldel.'_ysfor P.w_ves s_rrtplln__ locally_Lmucka_lF

_nisocropiccrust.

_ecause locale._r:hqu_ketraveltime residualshave no: been observed :o display!_0° periodicity,_d

the delaysfrom the NTS testsdo noc increasewit_ distancebeyond _bou: 70 kin,itisLikely:h_csources

Istruc:ures,stresses,subpaz:tile[cr:cks_nd micro_'rac:ures,or intrinsic_nistropy)responsib'_efor the

nuctenx_.esttravel:irneanomalies_re ac sh_ilowdaoths.The re_ion_[d_ec:ton of minimum horizontal

stress,in/erredfrom e_r:hqu_kefocalmech_nksms _nd in:._reementwith resultsirornR_inierN[esa.s_ress

measurements and from sh_i_.ow hydro(r:c ciecarmm;,ttons ac x.".'c:a, .L[oun:_.in, approxim.'_caiy coincides

'#i:h:he iow-veioc::y dSec::_n, _u_escln_ :h_c con_em.oor_r'! stresses m:ty be concribu:m_ _n a::mu_h_

imprint:o _cous:icveiocities,possibly_ccordtnz:_:_tee.x:ansive-d_ac_nc7anisotropymodel.

C:us:a_he:erozenei:.vprobablypl_ys,_l_r_jerrole:h_n f:=c:ure-inducedseismic_niso:ropytndecerminin_

:heobserved dei.'_ys.Rock _:shallowde,ocn(< S kin)in :he _[ientCAnyon c_der_,has towervelocityth_n

sur:'.oundin_country rocx,exoia_nin_in par: _he :r_.vei-_,imeretaxd_.tionfrom Yucc-tF!_c _xtdRa.hlier

.... ._,£es_:o northwestern SG_S:_" _:a:_ons, _:hough no such c_der_ is present :o the e_: of _'TS _o explain

:he e_:w_'d delay, Ex:mtn_ion ofcontour ptot_ofSGSffN stationdei._ysindicatesthac m_tny _r:tvel-

:line:,no.-.._P-'as_,re_eoZ:'=,_hic_:_y_.xed,_ndependen:oi :he auc_._.--devicerest_ourcere_ion,suggesting/

:ha: _oc:.t_cruc:uresr_therthan re_tona_stresses_re :he prirn_":/sourceso_ :he travel-time_nom_,Lies.

Where SG_Sh" cover_,geisdensest._,'oundYuc:a ._fount_tn(sLxst_.:ionsi,localvmri._:ions_re visiblein

veioc:::', f:r .,.x=:r.._[_. : s:ru::ur'.[ dLsc:n:i==::y '=e:ween C::ter _".:: _nd Yuc:'_ " ""

::c_:s ",'. _h0,Z_w _:a_. _,.'c;"u _icm..':e. h:ve =e,.'. _e:_r.-..::t:d :: -.::'e _'.,_:er '.'riot-:'," ::_..", c!_::: :_cks_c sh_ilow ciepchs _c XTS, Ei_h-veiocicy :r_ve_.th'ne con:ours :rend roughly pax_ilei_o _he CP _hrusc,

which exposes m_ny c._oon_e rocks.T:tus.app_en: _nmoc:'o_yin P.w_ve veioc"tym_,y be attributed

inp_r_ :o shallowstressor:en:acianrei._t_ve:odiscontinuities(fr:c:uresand f_,ui:s),Zthologicdi_erences

_c sa_Row depth (,._'_onaces versus c£_cic_}, or both. T:_e ::_lon_ network :r_veA-cimedet_ys mo,y

• _',erefora provide use_'ui _:orm_,t_on on :_e direr:bunion :.:_. or:enter:on oi _quders _nd _qu_caxds, in the_osense o_'_ther :homilies,

_.',

.., -- ,, : _ ,;_ _0 _ 0 D

_t 0 O_

'.J : :3

-'-. A 0 _"

•t.3 --

Y

,,,._au rn Ioe_J

=._._3,a#,,,.::'. -- .:.z:.:u:.'.a_. llsc -'=_-uc=:n "z_c_ -:=_n_J.es

',,!asa:asc is_to -'-'=."Cpan ¢,,=:_s _._:i_:a oVsar':ea inlays; sol- -

i.-._ica_o "un_:=on values.' ]4

,' )

Fracture Counts from Borehole Logs

by F. H. Nelson, R. Snyder, and J. E. KiblerU. S. Geological Survey, Denver

Fractures detected by the sonic waveform, televiewer, and television tools inboreholeH-4 at Yucca Mountain are plottedas a functionof depth alongsidecaliper,density,resistivity,and flow logs.Of the threefracturelogs,the sonicwaveform logprovidestheleastinformationon individualfracturesbecausespatialresolutionislowerand the fracturecount islessthan thatrecordedby the televieweror televisiontools,The televiewerprovidesa lowerfracturecountthan the televisiontooland almostallfracturesdetectedby the televiewerare alsorecordedby the televisiontool. Theteleviewer shows that most fractures dip steeply and have a median dip angle of 79degrees (dip angle could not be obtained from the television tool). The azimuth of thedip vector is roughly WNW. If a dip=measuring capability could be added to thetelevision tool, then it would be the preferred tool for recording fractures because morefractures are recorded with it than with the televiewer and because it operates in bothliquid-filled and air-filled boreholes (the televiewer requires a liquid-filled borehole).However, the television tool's image may be obscured if the borehole liquid contains

-excessive amounts of suspended solids.

The flow log from the H-4 borehole shows that the most permeable zones lie inthe upper parts of the Bullfrog and Tram Members of the Crater Flat Tuff. Thetelevisionlog shows thatthesetwo zones are among the most fracturedzones in thelower 650 m of the H-4 borehole. Resistivity and density logs indicate that neither zoneis altered (zeoIitic), Based on information from cored boreholes at Yucca Mountain, thelow resistivity zones are interpreted to be zeolitic with minor smectite: in general, theyare zones of no flow or low flow witl; few or no fractures. The lower part of the TramMember is a good example of a low resistivity, no-flow unit with few fractures.

The flow log was divided into 37 intervals and the change in percent flow vs.fracture count from television was plotted. The plot shows a number of zones withfractures but no flow, one zone with flow but no fractures, and a scatter of data thatshows no obvious increase of flow with increased fracture count. However, whencumulative flow vs. cumulative fracture count is plotted, then a monotonic step-wisecurve demonstrates that flow does, in general, increase with fracture count. In thisborehole, there are no prominent single-fracture "thief zones" contributing most of theflow nor are there zones of intense fracturing contributing most of the flow.

15

I I II III II IIII

F-_acture Stu'dies in the Welded Grouse Canyon Tuff:a'

Laser Drift of the G-Ttannel Underground Facility, Rainier Me_a,

• Nevada Test Site, Nevada

By S.F. Dieml, M.P. C,_mrmac,v, H.S. S_:Ifs, and J.K. _dum

We stuaie_ fractures in tr,e welcem Grouse Camvcm _uff in tee Laser Drif=

mf ire _-?unmel L_csrgr_una Faci!i:_/ (G,-LF) locate= !m Aaimier (_e_a, am:u: b4

km mcr_._w_st of _ercur_/, _evaca. ?he Grouse Canyon Tuff has litmalogi=

Ormcertim, stresn cmmditioms, area am ovmr:uraen :em'.m c:m_ara_le to trcse of

t_',e =r_:mmea rmm=s,.'.-.r_/ r,_r_.::n a_ Yuc:a Mountain.

=._m_eralhmrizontal _xoremales _ere cmrea in ire Laser Drift as :art mf the

Yuc=a Mcusntain _:ra_=t,/cetesting =r=gram. The L_G-'-c.-m_uc'.eas_r've,/s in tr_ge

- mu _.me_x=r_leg usi.-,g = =crer_le vi=_ came-a _ vicem tace re=.r=e, Vi=e= tamem

of tr_ :(:_emoles were usea to Ce_ec.-.the _resemce and orientation of fractures

lm_ersec:ing tlm Ccrer.oles. The mumcers of fractures omserve_ in tlm vide_

_ur_ve,.,s_.mmare we_l _i:m fracture c.-umts _.urzmg t,'eexamination atom Io_glmg af

t_e tee:were{: c=re samoies from tm_ _x:rerxo1_.m. The vicem tame rmc.=raing _id

:f_Ir t,,'_ aavamtage of a m_re ac=ura_e c._mt of fractures in ru=ole z-.rams, welch

were immQsIi_:II to rec=n_truct from t_,e c.-re cx_ce it maa =men renw:vea fr:m the

m:r_r_le.

The _x:r_olIS were criemtmK= to £n_irs_=t t_,e :r:mim_nt fracture tremas (N.

25" E. and N. 40" E.) e._sea in tee Lamer Drift. MaIt mf the frac=ureI arm

relativeiy 01anar with mear-ver=i=ai Oi:s. Minmraiiza=Amm ale:ragt_ frac:ur_

general iv cmmiistI of irmm and mangameIe staining, =u_ cane _re_,:minant fracOArm,

tr_moimg N. 5 E. -Ate a :in of 8a' .:E. is fi11_ with czar.

16

_:i_ca_.c=ms c._mmcmlv arm =mr=mmmZ='.,!ar am= mara_lel _m wel_ing a_c are acuno_

zmrr=ml,/ seal mZcr=grac_.urms, wmi=m £n_ica_.ms _rmc.!:£-.acicm =f _.-.ese minerals

-=m fluzcs nxov_ng _mrcugm _'e mi=r=frac'.urH. #_cularia anm £r=n, _anium,

amgameee, amc ram=. ear%,_ minmral =m,as_s _i!l _e microgram:urea. A fe_

:ular_.a-fi_:-=c mz=r_frac=urms arm. =arai_-ml :_ atom a_ aS' _ _.,_,e=iame of

:- i=img.

The mi:rmfracturem am=ear _m =a mx_mm_=nai in mrZg'.n. A _r=f,.!e cf _ru=-

=ressurem fr=m _,vCrauli= frac=urm ".ss='.m_in _ vert-'-'=almelee near the _,'LF

.'.m (War=in_i am= =_.-.ers, !_i) imci=a=ms =_a_ _._.m leas= mcriz=n_al stress- o.,

a_mitu=e in tr,e _elcec Grouse Canyon is ma_f am much as in tee amjacen:

:m_l_ec unZ:s. Thus, _e _!=ec uni_ '.sin a s_.a=e =f e,_%_mslcm, =rcma_Iv

Alt_.cugm _me Grouse Camvcn _uff is c:rrmm_.Iv im _,_e_msa:ura=_ .-=me, _re

_cr=us mimeral =mame_ _ma= =:a_ ._rac'.urm _ur_acms amo s_al mi=r_rac'._re_

-am _re m_=r=frac-:res _r_ =_=e ,_--=_=-mmec.'._K3. flea _a=ns for fiui:: m_vemem_.

ar=imsKi, N.R., Nmr%_rm=, D,A,, _Jcrmi::. R.A,, Vci_em=cr, _, 'A.C., a_C

F£n_e,/, S,J., i_Si, The _=rma_icm in_mr_,a=_ frac-ur'.mg e_=erimem=--an in

$itu imvem=Iga=i=n cf _V::rauli= frac=_re :emavimr near a material =r=:mrt,/i

£m=ereaee: Sam:ia National Lacorat=rimm Ne=mr: _:ANCBi-JgN, 82 _.

17

Dismbution of Fracture-liningminerals at YuccaMountainB. Carlos, D. Bish, S. Chipera

Fracture-_ing minerals in the tufts at Yucca M0untam, Nevada are being examinedas 'partof an ongoing study to characterize potential flow paths away from the proposedrepository horizon. To date, only an incomplete knowledge of the lateral variability infracture mineralogy has been obtained because of the limited number of existing coredholes and incomplete sampling of most of these cores. Portions of the Topopah SpringMember have been examined from USW G-2 (north of the proposed repository,block),USW G-t, USW GU-3 (south of the proposed repository block), L_25a#1 (east of theblock), and water well J-13 (in Jackass Flat). Only the core from USW G-4 has beenexamined over its entire depth, and fracture coatings in that core appear to be related tostratigraphic interval.

Tiva Canyon lVIemberSIost fractures in the Tiva Canyon Member of the Paintbrush Tuff have partial

coatings of m_ganese oxide dL.ndrites (10-60% coverage) and fine grained white powderor crusts of palygorsldte with minor modenite and opal CT (5-40% coverage), Rancieite isthe only manganese oxide mineral identified so far, but other minerals such as lithiophoriteor todorokite may be present. Many other fractures contain the silica minerals mdymite,cristobatite, and/or quartz coating 75-100% of the fractu_ surface. The manganese oxidedend.n,tes and fine-grained white silicates are generally on top of the silica minerals whenthey occur in thesame fracture.

T0_pah Spring MemberFracture coatings in the devitrified interior of the Topopah Spring Member of the

Paintbrush Tuff can be divided into three groups. The first _oup consists of thelithophysal fractures formed during devitrification of the tuff and formation of thelithophysai cavities, which they often connect. Like lithophysae, the fractures havebleached altered zones surrounding them and coatings of tridymite, or cristobalite or qu_zpseudomorphs of mdvmite. Although the fracture surfaces are now 100% coasted andmany are sealed, later deposition of calcite and other minerals demonstrates that some flowpaths remained open after the imtial formation of these fractures. The calcite commonlyfills the lower part of the lithophysal cavities. Zeolites occur with or without calcite inlithophysal cavities and brecciated tithophysal fractures in the lower lithophysal zone inUSW G-2, and euhe_ quartz occurs with or without fluorite over lithophysal coatings inUSW G-3.

The second group of fracture coatings consists of smooth, nearly planar fractureswith partial coatings of manganese dendrites and fine.grained mordenite. Lithiophonte isthe only manganese oxide identified to date. but scanning electron microscope examinationsuggests ranc:eite and/or todorokite may be present in some samples. These fracturescrosscut lithophysal fractures, but are believed to be early, possibly cooling fractures.Slickensides have developed on many of these fractures.

The third group includes all later, generally coarse-grained, fractur_ coatings. Latercoatings may occur in re-activated earlier fractures or on rougher later fractures. There isextreme variability,in mineralogy and distribution of these coatings across Yucca Mountain.The fracture-lining minerals are generally visibly euhedral, and the fractures are notslickensided. Coatings include heulandite, stellerite, and mordenite in USW G-2 and USWG-l, rosettes of smecrite clay in USW G-l, heulandite and mordenite in USW G-._andUE25a#1, and drusy qua.nz in J-13 and USW GU-3. In GU-3, flourite occurs with someof the quartz. Fractures may be partially sealed or open and probably have channelsresultingfromirregularitiesinthecoatings.

FracturecoatingsinthebasalvitrophymoftheTopopahSpringMemberarefine-grainedwhitetobeigewithmanganesedendritesandappearsimilarinallcores,butthe

18

, _ ii ii I IIII IIII IIIII

i

!,

mineralogy varies laterally and with depth. Fracture coatings cover I00% of the fracturesurface and slickcnsidcs are common. Smectite, heuiandite and manganese oxidespredominate, be erionite has been identified from sing!e samples in USW GU-3, USW G-4., and UE25a#I. Phillipsitc occurs in UE25a#1 a meter below the erionite. The presenceof different mincr'a.ls in close proximity suggestflocalized conditions and limited flowwith.in the vitrophyrc.

Fractures in e non.welded Topop_ Spring Member below the vitrophyre fromUSW G-4 and USW G-2 contain a dvase of clinoptilolite, sometimes overlying thecristobalitc. This interval has not been examined in other cores.

Taft of Calico HillsFew fr'acu.u'es occur in the tuff of Calico Hills in USW G-4. Fracture coatings

predominantly mordenite and clinoptilolite, covering 80-100% of the fracture surface.Because of the fibrous nature of the mordenite, fractures may be more permeable than therock mamx even if the fractures are closed.

Crater Flat TuffThere is a close relationship between matrix and fracture mineralogy throughout

most of the Crater Flat Tuff in USW G-4. Zeolites are similar to those seen in the tuff ofCalico Hills and occur primarily in fractures in zeolidc tuff. Coatings cover 100% of thefracture surface. A few fractures in zeolitic tuff contain manganese oxides and hematite,wluch tend to be diffused into the matrix.

Devitrified intervals have a few corroded lithophysa.l fractu_s. Manganese oxidecoatings are common in the dev_tm_ied intervals, with or without quartz or calcite. Qu__d calcite-filled fractures are often sealed. The manganese oxide minerals are mostlypyrolusite and cryptomelane-hollandite. In the Tram Member, manganese oxide mineralsare inter.oTOWnwith spherulites in the matrix adjacent to the fractures. Fracture surfacecoverage in 100% in almost all fract_.u,'es. Hematite occurs with manganese oxides in mayfracna'es in the lowest part of the Tram.

Below the water table, pump tests provide more information on hydrology thanf'racn.u'e mineralogy does. More than 75% of the water produced from USW G-..t camefrom the interval in the Tram Member containing a0undant manganese oxides and hematiteon fractures. As the water producing interv'als are different in other holes, and the corefrom those intervals has not been examined it is not known whether manganese oxides ar_generally abundant in water producing inter,,als in the Crater Flat Tuff.

19

STRESS _ AT YL'CCA MOU2qTAD;AND LNSUR.ROL,"NDLNGREGIONS

Joann x,[.ScockUSGS, MS 977, 345 .'vlidcUefieldRoad, Menlo P_rk CA 9,.t025: also at Dept. E',wth &Planetary Sciences, Ha.rva.rdUniv., 20 Oxford St., Cambridge .',,L-k02138

The stressfieldinsouthernNevadais¢onsa-zinedfromobservationsofearthquakefocalmechanisms,hydrauiicfracrunngstress"measurements,boreholebreakouts,anddrilling-inducedhydraulicfracruares_:.FocalmechanismsincLica:enormalfaukingonNE.smking planes, and smke-s!ip faulting on variably oriented planes, suggesting a combinedstrike-slip and normal faulting stress regime at seismogenic det)ms (5-15 ka'n),with the leastpnncipal stress, _, approximately horizontal and oriented .q'W-SE to E-W. Because bothstrike.slip and normal fauiUng mechanisms are observed, the maximum and intermediatepnncipai stresses, a'_ and a':, r_spectiveiy, may be close in magnitude to one another.Hvdrofrac tests at Rainier Mesa and borehole breakouts in _ holes at Yucca Fiat and Pa_ute,Mesa give S._(least horizon_ principals_ss) directions of N45QW-N55_)W andN,.tS'W.N60°W, respectively3:'t

Magnitudes of stresses within the saturated zone at Yucca Mountain were determined_om hvdrauiic fracturing tests in four weils tL'SW-G1, USW-G2, USW-G3, and Ue25-pl). InI2 ter,(s from 650-1700 m depth, "new' hydraulic fractures were created to determine themagnitude of S_. In all cases, me ver_cal stress, S,,, exceeded both the measured value of S)_and estimates of the maximum horizontal _nnc!p_ stress, S_. These measurements indicate a

normal fauitmg stress r_gime, wire values 6f <p=_cr,,-_)/(o'1-oh)ranging from 0.25 to 0.7. Theobserved values of Sh a.r_close to values at ,vhich',,:rictiona2sliding might be expected to takeptace on optimally oriented preeydstmg faults, suggesting that thestress regime may be nearfailure by extensional faulting. Drilling-induced hyck,'auiicfractures in three of the wells, andborehole"breakouts tn two of the wells, indicated an Sh direction of N60°W to NdSW, inagreement with other stress field indicators.

Three hydraulic fracturing measurements were attempted in the unsaturated zone inUSW-G2, but mese tests a.i.iappeared to have reopened pree.'castmgfi-actures and hence onlyplace upper bounds on the vaiue of or-,. An e!ast:c metier assuming lateral restraint has beenused to predict the ma_itudes and orientations of the principal stresses in the unsaturatedzone_, but these have not been coru'u=medby ac,"ualmeasurement.

Tae water level in me holes a_ Yucca blo_t_n ranged from 385 m to 752 m below thesurface, bleasured values of S_ were le':.sthansun'ace hydrostatic pressure (the pressure of a,.oiturm of water reaching the surface in the d,-,ilhere). Under these ck,'cums_ces, favorablyoriented water-filled frac_u.r_smight propagate ii the fluid pressure were to increase. This mayhave happened during _lling. in th.,-eeof these hotes, as cu'cuiation of d.riLIingfluid could notbe maintained back to the sun,ace and large volumes of dniling fluid were lost at depth. Longvertical frac,na'esvisible on the bor_hote re!erie'.vet iogs from mese holes are interpreted to behydraulic tracrares formed during d.riLling. These drifting induced hydraulic fractures, inconjunction with the low S), magnitudes, are important because they suggest a dynamicbalance between the magnitude of S)_and the height of the water table. Thus, a substantial risein the water table in the future might lead to the opening of pre-existing fractures, or thepropagation of new hyd.rauiic fractures, there:by dra.stically affecting the saturated zonehydrology at Yucca Mountain.

(I) Stock et al., _ _ IL¢,_v. _ 8691-8706. 1985.O '=) t'3'_ •(2) Stock and Heaiy, in _ _ _ pp. o_-:,.., 1988

(.3)Springer and Thorpe, Rep. UCRL-87018, LL,'¢L,1981.(4.)Warren and Smith. _ _ _ v, _ 6829-6839, 1985.(5) Swol.fset al., in __ 1_790.pp. 95-101, 1988.

[for CASY conference, September 13-1a., 1990]

2O

FIELD DESCRIPTION AHD HEASUREHENT OF HESOSCOPIC FAULTS WIT}{APPLICATION TO KINEMATIC liND PALEOSTRESS AANALYSES

8

Sco_ A. HtnorU.S. Oeolof_cal Surve.v,P,O, Box ;_5046, H5 9_3,

Denver, CO 80225

As w1:._ other types o_ fractures, faults can have considerableinfluence on the flow of fround water, actint either a| flewcondults or barriers, or both dependln| on spatial variationl infault characteristics. Physical characterizatlon of faultl canbe useful to hydrolo_ists in developing realistic Around-waterflow moceis of faulted terranes suc_ as that at YUCCA Hountain,_evada. Perhaps less apparent is the utili_? ot studiesaddress_n_ the kinematics and causes of lausanne, AncludAn|paieos:ress determina_'.ons. Knowledge ot these aspects offault:n_ _n areas ot _ood access_bAIi_y can freatly faciIAta¢epredic::ons o¢" fAul_ geometries and An_ernal fault structure innearb.v subsurface rock masses lacking Adequate structuralcontrol.. Furthermore, derivin_ orientations and es_aat_n|magnitudes of paleostresses associated with various fault,s|episodes can be valuable in predicting how the same frac¢uredrocks w_!l respond to the present regional stress field And tolocal, man-znaucec s_ress chanles resul_Ang from repositoryexcavation. In the Yucca HountaAn region, mesoacop¢c faults --_'aul:s tha_ commonly can be seen _n their ent_retT _n outcrop and_ha_ _enera!l._ show net offsets of < 5 m -- are best exposed and!end _ne._seives _eAi to detailed ooserva_Zon.

Heasurabie aspects of mesoacop_c faults include: _I orientat¢on;) shape (i.e. deviation _rom planar geometry) ; j ) surface

dimensions; _) ra_e o_ slickens_de striae; _) ne_ oftse_ lotsepara_:onl; _) :'sulk-zone w_c:h; 7) aperture; and 8) faultspacing. O_her _mpor_an_ elements o_" fault zones are compol_onand s_ructural :'Abr_c. Fault zones u,ually _ncAude variouscombZnat,.ons and arrangements ot taul_ breccia, clay gouge, And(or) subsidiary, fractures, AncAudAn¢ Riedel s_ears and tenl¢cnfractures, Critical in _inema_c s_udies As the determination sIPslip sense on individual faults using: I) isometrical relationsof various _ypes of subsidiary fractures Ln and bordering thefault .,one, and gene_icaliy rsla_ed to L_; 21 asymaetr¢caApolish; S} _ool mar_s; 4) vesicle smears; _) arrangement of 8yn-slip mZneral growths; 8) dr_ folds; and (or) ?) o_fset featuressuch as marker beds. I_portan_ observaclons concernAn| ¢herelative a_e of faul_s are: I) presence, _r_e, And spatialdistribution Of pre-, s_n-, and pos_-sllp minerals; 2) cross-cuctin_ relations of faults and sllckenslde striae; and 3) ageconstraints from faulted s_ra_igraphAc units. Attitudes o!bedding, compaction follations, or o_her paleoda_ums anddetermination of paAeoma_net¢c directions in _auA_ blocks arenecessar? to test _or horizontal- and vercicaA-ax_s rotations offaults, respectively.

21

tauAt-siAp data, which consist of fault and slickenside-striaeor_emtat_ons and s!ip-sense determinations from one or more fieldaltos0 _an be qualitatlvel7 characterized usin_ equal-area androle plots. Fault data su_dlvided with the aid of theseIripn:cal plots can be inverted usin_ established computationalaet.edj to fLnd best-fit orientations of the principal stresses

_ _, and am, where a1_s_). The computed ratio z=(a 2-e_/!e_-e 3) _s indicative o_ t_e relative magnitudes of theprLne_pai st=esees, Extreme vaiues o_ the ratio, which rangesfrom 0 to I, An conjunction wi_h o_her analytical parameters,LnaLoates _hat two 0_ the th=ee principal stresses can not bereLLaoiy dis_in|uished from each other usin_ the analMzed data#uOlet, A cemputatlon-intensive iteratlve clusterin_ technique_an be used to _urthe= separate from a mixed-faul_ data set two_r _or_ |u_set8 that are compatible _ith unique s_ress solutionsan_ are ccneAs=ant with known rela_ive-a_e information. Through_Aa _rOCeal _he _aultln_ and _aleostress histories of an areacan _e _ore clearl7 established than is possible throughconventional =ethods.

22

RI-_OLITE FLOW FIELDS AS BARRIERS TO PAI.EOHYDROLOGIC FLOW IN F__STERN _'IEVADA

AaN'I_WESTERN UTAH BETWEEN I..ATITU'DF.S37"2,0' and 38°N.t, t

by _.. Ernest Anderson and Theodore P. Barnhard

Spectacalarly straight north-flowing drainages with ld_ly symmetrical cross prof'des are developed on

Mi_ene basia-t'dl strata composed of silidc volcanic dasts within a 100 km a area, iaformaUy known as Eccl_

basra in easternmost Nevada, and a 12.km a area along Wide HoLlowwest of Enterpmc in adjacent westernmost

Ut_'. In Eccles basin, a separate set of slightly less straight northeast-flowhag drainages with strongly asymmetric

cross profdes are adjacent to and, in part, overlap the north-flowing drainages. A similar set of strongly

asymmetric northeast-trendi-g d.rai,ages is present haa 2.00kzaz area hasouthernmost Hamlin Valley hawestern

Utaii. In _ areas, drainages lack the preferred orientation, strai_tnes.s, or characteristic cross profiles where

they_raverse adjacent or maderl.vingpre-basha-/'d.Irocks. Those rocks, prindp_y silleic Miocene volcame..s,are..,, -..

commonly fau/ted, tilted, and erosion_y beveled beneath the basin-fill strata. In places, the spectacularly straight

drainages are developed on erosional remnants of basha-f_ strata that only form a t_ veneer atop the deformed

SIioce-.v.e,.'oicarAas. In contrast to d::c,,'olcauics, the basLa-_ dasdc strata are cut by sparse ste-"o mostly nor:_-

or _rtheast-smki'_g faults and arc flat Iyi,agto ge,',dy tilted.

The drainage-pattern developme,_t must be contro_/ed by factors, such as contrasts in degree of

cera_=tatiom that involvelarge percentages of the basln-_ strata. Althouab identification of controlling factors

is bin:spored by poor exposures, study of sparse exposures of the basha.f'rlland lag debris in Ecrdes basin su_ests

that-the sbou/ders of the north- and northeast-trending i_terfluves are tmderiaia by steep drainage-parallel

pands withinwhlc.bvein-.typecarbonateanddisseminatedcarbonatecementissigni_ntlymoreabundantthan

interveningpartsofthebasin-F_dlstrata.Apparentlythesepanelsaremore resistantto erosionthanthe

uncementedtoweaklycementedsedimentsoftheinterveningareaswhichaxe,accordingly,etchedoutinto

straightchannelwa.vsduringthelatestcycleoferosion._ llthologiccontrolofdrainage-patterndevelopmenE

is,inturn,structurallycontrolledbysparsemap-scaJeandsmallercltainnge.parnllelsteepfaultsandfractures

Thec,_.rbonatewasintroducedfrombelowintothebasin-fallstratabycizculationinanancientfxactttre.controlled

23

_ound-water system. A veryhighd_wee of uniformi_ in ph_io_aphic expression in the subject arc_ sugg_.s

urdform average amounts of introduced carbonate such as might be cxpeaed in predpication from a deeply,'/ i

circulating regional ground-water system..¢

In Eccles basin and Hmnlin Valley the conspicuous linear drainage patterns give way northward to

normal dcnd.dr.icpatterns in equivalent basin.t'dl strata. In Eccles basin the boundary,between the contraar.ing

drainage parterre is abrupt and trends e_t-w_r.. The b_in.fill strata on both sides of this boundary, have a

uniform silidc volcanic clast a._emblage and appear to have similaraverage concenr.rafiom of carbonam c=mear,.

Apparently the chief difference is in the d_m"outlon of the ctmentiag material; the dism_bur.ionof carbonate is

uneven and fracture controlled in the topographically high sou_ern parrs of the are= and relatively uniform in

the lower northern parts. We interpret t2m difference as a reflect/on of paleohydrolog/c conditions relamd to

i) ponding and rising of shallow elements ofa south-flowiag regional carbonate aquifer sys[em as ic encountered

Cenozoic rhyolite flow fields as aqukarda, and 2) incomplete mixing of the fracture.controlled rising carbonate

waters with shallow loc,rdrecharge from the siliceous igneo,_ rocksof the Clover and southerumosc Indian Peak

raneeslocatedsouthof[heareas ofbasin-t'dls_ata.

24

Sa=u:aumd-Za_ G=ound-WaEm_ rl_ au Yu==a 8ountaln, _evacUt; Can FEa:=_e FI_

Be _equaualy Cha4:__i=md2

SO_',nB. C=arnec_ and _._u: L. Geldon " U.S. G4aloqiEal SuF1ev,

Lakewood, C=IaEa_o 8022".

TP,m hydE¢lOc/y of _uo=a Hounta.L.r., NeTadA, is betnq EP.a.cacT.eEi=ed by _e U.5.

GeolaqL:al Su=vey co evaluate L:s ,ulca_ll£_7 as a cease,co=7 _=: high-level

nu:Ima: wastm• C._u&_e=i:a=ion 0_ _=e satuE&ted-z¢ne flow ayet_ o_ _u::&

Hountaln and vLoini:7 p::v_de| ,u_:Jtant_al ev_den:m _Pmt, &: =he s:als a% :he

.-mpEmi=or/, .a¢.=Ees have a Za_qe ed.'e¢¢ on gE:und-wauaE _Iow This evLdenc:s

L._oludee u:ace_e¢t:: and tem_eEa:u:e muEveym £n unca|e_ d:illh=lee =._a: snow

ml

..:w o¢=u:::._9 L._ _::en=Le8 a: f:a¢':uu:e Loca::.one (de=s_ned .'.-:m a:ousti:

:elevimwea: Logs). Addt:ional ev_dsn¢e iJ _ndl:ated f:=m hyda:aul£a.:eat_.nq

=esulUs :._at y_sld :._:ee-<:::_¢nent ¢=awd_n cu=vee c=::ss_ond_nq :: =a/ease _

waua: _::m: (1) _=a==UEeS only; (2) :_e ::¢_ :at.-:.x: ted (3) bot_ _=a:tu=es

and _he ::ok :a:=_¢. C:=_au:ljon :! da:a "--- eleven _-'' '_'-'--...... ,.,...am at Y_¢.-&

Mountain shows :w&: wate:-_Eoduc:._q isis=vale oc:'.:: :_-a::ug_out :he Te_'uLa._ i

8t:atiq:a_Lo o:Lu=:, and :._a: r.o ainqLe _o==at!on _= :am=e: ;=:d_¢es we=mE in

all beEch:Lee. _tsa:-pE:du:_nq 5ntmEVals a_au:sntly co:us: as & :saul: o_ :_e

_:=:uiuous £nte:sect_=n o_ waceE-:eau:a.nq _¢U:ta, s_eau:s, and :eel:s, wni:._ a_e

not :elaued :: :._e _eqEee o_ weldinq m: :_s dog:as o_ !=ZmtUE:._q. FOE

example, a: :W.e C-holm c::plax, a 1,027 mquau:e :mimE, =el:A-well si:e _u:_ut 2

kiloumtmEs sou=Sweet o_ :_e dms_qn ::e_=s_:o:7 sa:ee, :._s pELnm_pal wets:

_==du:_.nq zone in one well was in :::welded so pa_'_=Lally welde_ ourS/ ::8: o_

:_a watt= pEudu:':&on tna sec:nd well was tn :_der&tely co: densely welded

:u_; and ware= in a "-,2Lt=d well was pE:dume_ auJ=equAlly _==m nan-welded _:

pa--=ially welded =uf_, :ode=atmly :m densely welded =u_, and :u=_ b=emmia.

Aln_ouq_ aouu_-eout_s&s:sEly := aou:_-s:ut_westaa:ly ==sndinq _:a_=ua:s|

_Eedm_.ate Ln _._m C-h:lee, no 8inqle we: o_ _Ea=:Ua:SS is assoc_a:ed w_:_

flu£d inflows o: pEa_um=_=n zones £dencCf_sd by :_ez&tuEe loq8 an#

::a:e_e_==: au.-'N_e _ :_emm wells•

25

26

ModelingofFlowandTransportinFractureNetworks

by

KentiKarasakl,Amy Davey,JohnPeterson,Kevin Htsrir, and Jant Long

Fieldevidenceof_qtctunscontrolledflowinhardrocksincludesa IAckofhydrolosic

connectionsbetweenboreholesandklshlylocaliz_tandf_c_.mssoclatedheterolene-ous transmissivities in boreholes, Di.f_cuities ust:_ated with i.nt_retation of fieldmeasurements in such hetemlleneous systems and making of model inputs are discussed.The problem of scalint up is one of the most important issues that need to be addressedwhenconstruc_llaregionalscalemodelusingsmallerscalemeasurements.Assuming

the flow system as havinj a _actal nature may prove very useful, Porous mediummodels and fractme networkmodels Lrecompeted and the llmttsaons of _th appruachesare discussed. The laritest d.l.ffemncebetween the two approaches ties in the inputgeometry rather than in the f_ndamental numea'icaltechniques, The importance oftheconceptual model and the philosophy in the apptication of numerical models isemphasized. Recent developments in model,Insof flow and transport in f_cture systemsare introduced. An adveccion.dlspersion code that uses triced Eulerian-Lasranlian,adaptive llriddin8 tech_.iqueis outlhqed. The model m!n!m!+esthe numerical dispersioneven under ktllh Peclet number conditions, An equivalent disconmauum model and newinversion tecl_ques arc introduced. The model does not attempt to reproduce everylleomcmcal detail of the real system. I,nstead+it atlempts to reproduce the observedbehavior of the fractu_ system usinl simpU_ed 8eometrywhile preservin8 the system'sinherent discontinuous name, The inversion method uses an aljorithm called simulatedatmealhal which employs a statasticalrelation to pe_orm a random Ilobal search for thefracture network that best describes the system behavior, This search is equivalent tomakinl a simulation of the _cmre network, conditioned on hydrologic measurements ofthe network. By chostnll d_emnt seeds for the pseudo.nmdom number Iieneratorwhichdrives the search, one can also use the method to obtain con_ence intervals for precttc.

don of the hydmlolrlCprop_es of the _'acmru system. In this way it is possible to quan-d,._the uncertainty in the hyd,rololicpropertiesof a _cmre network. An example use ofan interamd function system model, that is applied usinl the put UE-25c-hole hyd,raullctest data, is also shown. The model is Ileneratedbyasetof_tt_e transformationswhose

coetqlclents_,'edetenmnedbyanopami_tiontechnique.

27

Fluid Flow in Rough.Walled Fractures

Robert W, Zimmerman, Sunil Kumar, and Gudmundur S, Bodvarsson,t

SciencesDivision

LawrenceBerkeleyLa_ntoryUniversityof C_ifomia

Berkeley, CA 94720

Yucca Mounca,i,n,Nevada,the proposedsite of an undergroundr_cLioacdvewaste

repository.,is composedmainly of volcan.tctufts, someof whichare hiffhly fPacuz.red.The hydzaulicconducUvitiesof the h_cr._d formationsat YuccaMountainorecon-trolled to a large extent by the conducuvitiesof the incLividualh'acmtes. As it is_fflcuit to measure the permeability of a single _c_..re in suu, it would be adva.nta.geousto havea methodof reiaUnl the permeabilityto more easilymeasuredproper.ties,suchas the f_ctu,reroullhnessprofile,percenut|econtactarea,etc. We haveusedv_ous mathematical models of rock _tc_..res to study the effect of these physic_

pard,meterson the permeability,

_e effect of the roulhnessof the fracture sue'acehu !'_en studiedusing a tsinusoiclalmodelor"theapenarevm'iation. At low Reynoldsnumbers,flow in suchafncmre is governedby the Reynoldsequation. We solvethisequationexactly for the_.'o casesof flow parallel to and transverseto the sinusoidalv_acions, and then usethe leomemc mean of these two values to esn.mate the over_l permeability, Wethereby ornve at an expression for me hyd,_ulic aperrate in terms of the mean and thest_dard deletion of the apertu_dismbution.The res_u ate in closeagreementwiththe numericalvaluescomputedby Brown (1.Geophys.Res., 1989)and Pa_ and

Chenil (1, Lub, Tech., 1978). We haveven_ that the the results are affectedonlyslitlhclyby the additionof addition_ sinusoidalcomponentsto the aperturedismbu.taon.

A ci_ferentmodelhu beenusedto use, the effectof contactareason theper-

meability. _ this model, the _cnu'e is consideredto consistof two fiat, parallelw_s, proppedopenbycyil,n_calasperities.A Bn_.type equation,which isa

hybridof the Navier.StokesandD_n:yequaUons,isusedto modelflow d',rouilhthef_cnu-e. This equa_on accounts for the viscous dral a,lonll the sides of the asperities,which hu been ignored In previous models. A closed,form expression is derived forthe permeability in tm of'the percentale contact area, and h/a, the ratio of the aper-nu'e to the uperity r'acLtus,The resultsshow that for contact _u on the orderof 20.30_, the uperides can reduce the permeat_iUtyby as much u 50% below the

p_ei-platev_ue h2/IL

USO$/CASY "Fractures,Hydrololy,andYucca,V{oun_" Symposium(13.!4 Sept. 1990:Denver)

28.......................................................... --

i

HodelAnl o_ F!ow an_ Solute Trans;cr_ _hroul, a Vat,isle

Aperture Par:_allySa_ura:eaPrao_ure

bY

Richard W, HeIAp and [_war:H. KwIQ_IAs

U._. G,o_¢tXQli Survey, :enver, CO

nu_erlcai moae|. _,u aeen Ceve'._:e_. far sL',uAa:'.nl water

_n_ solutemovement :nroucna single rcc_ frac'.ure,The _'r_c;ure

Is :_ncel}tuali=e(t as a _wo-cllmensAonal grid oi' nodes, Al=er_ures

vat? fromnccie=c note, produce.hia fro=sureprci'l_,e_na--.resemolesi

wif.r_e Lron. Rancomvalues for :r,e '.:|.ncr,_lL!y dl.s:r_bute_ anct spitliAly

correla=eaaper:urelare leneri_eaby _onte CarLo s_muia_:cn,wi_b mean

_'h* ""'" _.lW'S _se'= ;0at,=v;r:ance (';AR)s:e_::le=.w,/:r,euser, ,,....... .

_liculate_ransmisslv_:7a_ earn node, Hys=eresLs In _he saturatloncurve

_.s_ncorpcrate_ln_: _e model f:r par:'.IL_,7sa:ura:e_, Solute _ranspcr;

;s s:=uii:edby par:'.' "'"",,...e :rac_Lnl..... .,.L_n:s no_ Lnc_.u=e<_in _,',e

,_cde_,rather _ne s_reaa oi"par_:cle :ravel_Imes _s uses as a measure ct

_is_ersicn,

A sensi_Avt._'t test was ccnaucte_'_:_ _e madei _o

i_Lus_rate_e _n_'.uencect VAR on fL:want _ranspcr:propertiesof a

_rac:ure, Team resuA;ssnowed :nat as VAR was Ancrease_,:be averaie f:cw

ra_e _,rcuin_he trac=uredecreaseswnA_.e_be varianceot _ne tZcw race

/ncrease(i.Average _raveA _e of _e par_ic'esdecreased,bu_ _e

_ravei-_imevariance increase_. Hysterealsin _e scours:ioncurve ai_o

lncreued as VAR was Increase(i,This morea s_c}ui__e useful in evaiua_!n|

29

30

,.... , ,,,,,,i,ii ,,ii IIll IIIIIIIIIIlllllllJlIIIIIII IIIII ---

Flow Through Variable Aperture Frsotures_

john L. Sm.ootPacific Northwest Laboratory

P.O. Box 999, MS K8-77Richland, WA 99352

ABSTRACT

The rote of fractures in unsaturated flow througi_ welded andnonwetCea _uffaceous rocks is a funaamental question being a¢_ressecl for;erformance assessment activities in the Yucca Mountain Project. As partof the U,S. De0artment of Energy's (DOE) PACE-90 exercises, a test casewas _vised to investigate fracture-matrix interactions for a meter-scale block of ToDopan S_ring welded tuff (TSw), Documentation of flowat small scales will provide input to Cevelopment of performanceassessment techniques which must reaiis:ically be geared to largerscales. The PAC3-90 proolem consists of modeling flow through a

et*vertical frac,.re in the TSw block with apertures specified to averagefrom 0.1 to 0.5 am, which are among the larger apertures exoec'ed tcoccur at ae_¢n in TSw. Fiow t_rougn the frac:ureo bioc_ was simula_eowith the PORFLO-3 code for nonnysteretic, one-phase, unsaturatedconditions, Results of initial simulation of simple aperture constric'ionsanO expansions incic:_te that the presence of variable a:erture fracturestendea to l::erturOa preoominan:ly aownwara flow field w=thin several cmof the fracture, A clara set generated _y laser scan of an ac*,ualfrac:urewas usecl to generate a fully three-dimensional moore!of the block. Grid-mesh size considerations did not allow for aclecluateuse of the completedetail in the laser scan so the proDlem was simplifle_. The resultsindicate that dry conctittonswithin the fracture induce flow from thematrix into the fracture; pulses of water =nput to the top of the frac',ureproduce a gradient of flow into the matrix,

dJL._. I _ IJ__ J ..... I ..... III _[I

_This work was supponecl by the U.S, Department of Energy unclerContrac_DE.AC08-TBRLO 1830.

31

VoidStructureandFlowin.SingleFracturesd'

L.R.M','er

E:mhSciencesDivision

Law_nce BerkeleyLabonttoryBerkeley, CA 9472_

_'_edevelopmentofa successr'ulgeologicrepositoryina _ct'uredrockmass

requires_ understandingoz"thefluid flowin the_racruresasarune:ionof sn'ess.

Testsconductedonnatu.r_h'uc_s haveshownthatsinIlephue flowdec_ues

muchmorenpiRlywithstressthanwouldbepredictedbyamodelrepresentingtheh'uc.

n.u'eastwop:u'alleiplates.However,aftercvclLng,mechanicalmeuuz'ementsi.n_cate

thath'_ctu_edet'ormadon,at[e_tinsomerockry'pes,islargelyelutic.To beginto

unders_'a.ndthisbehaviora Liquidmen _Vood'smetal)injecuontechniquehasbeen

developedinordertoobta,Lnc_tsofthevoidspaceinah_ct'u_underd_erentnormal

loads.Thesemeasu,rementsshowthatwhilela.rgevoidsrem_nevenathighstresslevels,

theconnec_onsbetweenthesevoidsbecomemore_ndmoretortuous.One approachto

modellingsinglephaseflowistouseascr_tiEedpercolationmodeltogenerateacorre-

latedaperv._ patternsirnLl_rin appeozanceto the Wood'smetalcasts. Frac_'e defor-

mationismodelledasauniformreductionina.[lape--s suchthatall rock:deformation

isaccorr,.mod_tedbyachangeinvoidvolume.Usinganetworkmodel,flowisco.[culated.

at eachsu'esslevel. Resultsshowthat this approachsarLsfacto_ysimulatestheobserved

relationshipbetweenaperturech_ite and flow.

Asanalternativeapproach,theareu of contactin a fYacmrehavebeenmodelledby/

a "c:u'pet"of cones,sothatthecontactareaincreasesasfract_t.n_det'om,.adonincre_es.

Assurr_ngh.n..cm.n_deformationisaccommodatedbya changeinvoidvolume,andusing

ane_'ectiveconductancetoaccountforincre_edtorruosiryduetothecontactarea,llood

32

_eemont withobservedresultshasbeenobtained.

Two phaseflow _d capiUa_pressurechzracm,"iscicsin a h'_co.u'eundersu'esshave

beensimulatedusxngshestratifiedpercolationmodel.As a zerochorderapproximation#.

rel_tivepermeabilityofthenon-wettingphaseisbasedon theflowthroughthecritical

neckof thepattern,r_atis.thesmatlesraperture_ong theconnecmdpathofhighest

apertures.Flowofthewettingphaseisbasedonflowrb.roughthecriticalconnectionon

oneormorepercolatingpaths,andincludesascalingcorrectiontoaccountfortormosi_.

Resultssuggestacriticalsaturationisneededfortwophaseflowtoexist.Lnthelabora-

tory,mercuryporosime=yme_u.n_mentshavebeenperformedona singlenaturalfnc-

ruretoinvescigamitscapillazypressurechorzcmriscicsasa functionofappliedsu'ess.In

_dc_rion.flowmeasurementshavebeenmade usingmercuryasa non-we_ingfluid.

Resultssuggestthatrelativepermeability(ofthenon.wettingphase)ismostsensitiveto

ch_gesinstressatlowstresslevels.Comporisonwithmodellingresultsin,camsthat

ther_l:tcionshipsbetweenapplieds=essandbothc_pillarypressttmch_racm,-iscicsand

relativepermeability0resensitivetotheformofaperture_scribudon.

33

,, Ili,, III IIIIIIIIII IIIH I

DenverFederal CenterDenver,CO 80225

C. Barton CUSGS-D)MS 913, Box 25046 L. Myer (LBL)DenverFedera.lCenter I CyclotronRd.Denver, CO 80225 Bldg 50A, Rm 1127

Berkeley, CA 94720B. Carlos (LANL) 4i5-486.6456ESS- I, MS D462Los Alamos, NM 87545 C. Neuzil ('USGS-R)FTS 843-6879 MS 431

NationalCenterJ. Czarnecki ('USGS-D) Reston,VAMS 421, Box 25046Denver FederalCenter S. Rojstaczer (Duke)Denver, CO 80225 Dept.of Geology

Old ChemistryBldg.T. Doe (Golder Assoc.) Durham, NC 277064104 148thAve.NE 919-684-5843Redmond,WA206-883-0777 J. Rudnicki (Northwestern)

Dept.ofCivilEn_.neeriagD.Galloway('USGS-S) NorthwesternUniversity2800CottageWay,Rm 2234 Evanston,]260208-3109Sacramento,CA 95825 708-491-341191_-978-4648

J. Smoot (PNWL)D.Gillies('USGS-D) PO Box999MS 421,Box25046 Richland,WA 99352DenverFederalCenter 509-376-1352Denver,CO 80225203-236-0546 .LStock-(Harvard)

20 Oxford St,R.Healy (USGS-D) C_-nbridge, MA 02138MS 413, Box 25046Denver FederalCenter E. Verbeek ('USGS-D)Denver, CO 80225 MS 913, Box 250,_6

DenverFederalCenterK.Kara.saki(LBL) Denver,CO 80225#ICyclotronRd.,Bldg50F 303-236-1277Berkeley,CA 94720FTS 451-6759 R. Wheeler (USGS-D)

MS 966,Box 2._046S. Ke!ka,r('LANL) DenverFed_a.ICenterMS D_3 Denver, CO 80225LosAlamos,NM 87545 303-236-1592505-667..4318

E. Majer (LBL)Bldg 50EBerkeley,CA 94720

S. Minor(USGS-D)P,O, Box 25046

34

,,, II I fill llll lllll l

Poster preseqtersE.Anderson(USGS-D) P.Nelson(USGS-D)MS 966, Box 25046 MS 96d, l_ox 25046DenverFederal Center : DenverFederalCenterDenver,CO 80225 Denver,CO 80225303-236-1584

J.Odum CUSGS-D)G. Bodva.,'sson(LBL) MS 966, Box 25046

DenverFederalCenterT.Brocher(USGS-M) Denver,CO 80225

415.329-4737 R.Snyder(USGS-D)303-236-1263

M. Chornack(USOS-D)MS 421, Box 25046 H.SwoLfs(_SG$-D)DenverFed_ Center MS 966,Box25046Denver,CO 80225 DenverFederalCenter303-236-5180 Denver,CO 8022._

S. Diehl CUSGS-D) TurnerCUSGS-D)MS 966, Box 25046DenverFed_ Center R.Zlellnski(USOS-D)Denver,CO 80225 MS 424,Box2_046303-236-1635 DenverFed_'alCenter

Denver,CO 80225T.'Do¢ ((]olderAssoc.) 303-236-47194104148thAve.NERedmond,WA R.ZlmmermantI.BL)206-883-0777 MS50E

Berkeley, CA 94720]. Downey (USGS-D) 415-486-7106

E. Ervin CUSGS-D)MS 421,Box 25046DenverFederalCenterDenver,CO 80225

D. Galloway (USGS-S)2800CottageWay, R.m2234Sacramento,CA 95825916-978-4648

S,Ha.,'msenCUSGS-D)MS 966, Box250_DenverFederalCenterDenver,CO 80225

L Kibler CUSGS-D)

K. Kolm (USOS-D/CSM)

S. Kumar _.BL)

W. Mooney (USGS-M)35

rarttdn.nt_ I. Coe (USGS-D)C ^_ MS9O9U$_C, AC_ DenverFedez'_CenterMS P.314 : Denver, CO 80225W_nltlon DC 2055 303-236-0914

492.8371L Cole (USGS-D)

A. BQch(CSM) MS 913, Box 25046Denver Federa/Center

S,Be_ (Buret-D) Denver, CO 8022.5

F_ Center N. Coleman (NR.C)Onvf. CO 80225 301-492-0530ao3.236.4t77

C.Cope.('USGS-D)T.BJ_ltlMt(DOE-LV) 303-2365177US__7_.794-7590 P.Covington(SAIC-D)

303-279-7242

C. BouihtonCUSG$-D)T. BradyC45C}$D) R. Craig(USGS-LV)T, Brother C,JSC}S-M)

I. Devine (USGS-R)C. Boulhton (USC]S-D)303-236-U_ D. Dobson(LTSDOE-LV)

702-794-7940C,ButeDenverFed_ Cenm' P, Domenico (USN'WTR.B)Denver,CO 80225 ArLington,VA

T.Buono6USGS-LV) W. Dudley(USGS-D)702-794-7088 MS 425,Box 25046

Denver,CO 80225

F, Syem. (USGS-D)303436 5606 C.Faunt(CSM/USGS) '

W. Can' 0JSOS,Sandia) C.Fridrich(USDOE-LV)11343 W. 38m Ave, 702-794-7587Whea_Udlte,CO303.421.6410 K. Futa ('USGS-D)

MS 963W, Caulseau(USGS-D) DenverFedora/CenterMS 421, Box 25046 Denver,CO 80225DenverFedera/Cenm"Denver,CO 80225 A. Geldon (USGS-D)303-236-5307 303-236-0854

D, Chesnut(LLNL) I. GemmeU(USGS-D)4[ 5-423-5053 303-236-5014

M. Clemik (USOS-D, '_ R. Getzen (USGS-M)303-236-1914 MS 496

345 Middlefield Rd.MenloPark, CA

36

K.Levy(Weston)L Gibbons(ARA) 202-646-67794300SanMatcoBlvd.Suite A220 ' S. Levy (LANL)Albuquerque, NM FTS 843-9504

J'. Gomberg (USGS-D) B. Lewis (USGS-D)MS 966, Box 25046Denver Federal Center D. Lobmeyer (USGS-D)Denver, CO 80225 303-236-6227303-236-4410

R.Lucky(USGS-D)L. Hayes (USGS-D) 303-236-5033FTS 776-0516

F. Maestas(ARA)D. Hoxie (USGS-D) 4300San Mateo Blvd.MS 425 Suite A220Denver Federal Center Albuquerque, NMDenver, CO 80225303-236-5019 S. Mahan (USGS-D)

P. lustus (USNRC) B. Marshall (USGS-D)USNRC, MS 4H-3 MS 963Wasttington DC 20555 DenverFederal Center

Denver, CO 80225I_..-Kersch(SAIC)702-794-7620 S. Manet CLBL)

Earth Sciences DivisionL. Kniper (I.,ANLRJSGS) 415-486-5837505-667-1918

K. McConnell (NRC).1".Krasou (GeoExp. Ind. Inc.) USNRC, MS 4H-35701 E. Evans Ave. WashingtonDC 20555Denver, CO 80222 FTS 492-0532 ,

303-759-2746T. McKee (USGS-M)

I. Kume (USGS-D)MS 421, Bldg 53 M.McKeown (Burec-D)Denver Federal Center 303-278-3908Denver, CO 80225303-236-5015 M. Mifflin (State of Nevada)

2700 SunsetE. Kwicldis (USGS-D) Las Vegas, NV 89120303-236-6228

D. MuhsR. Laczniak (USGS-LV) MS 424, Box 25046702-295-0136 Denver Federal Center

Denver, CO 80225G. LeCain (USGS-D)303-236-5020 C. Newberry (USDOE)

: FTS 544-7942L. Lehman (Stateof NV)

_- 612-894-0357 G. O'Brien (USGS-D)MS 421, Box 25046

37

Denvm'_'ederv.lCenter M.SchmidtDenva:,CO 80225 303-431-8550

G. Pam_'son (USGS.D) _ V. Schneider (USGS-R)303-236-5179 I_IS414

Reston,VA 22092Z.Permian(USGS-D)Bldg.RIB, MS 963 G. Severson CUSOS-D)Denv_]Fed_al Center 303-236-5032Deny=, CO 80225303-23_7886 G. Shideler (USGS-D)

303-236-1273c. Pcm CSGS.D)303-_ I_Shmba (USGS-D)

MS913,Box 25046D. Plcmf'fCUSGS-M) DenverFederalCenter345 _efieldRd. Denver, CO 80225Menlo_ark,CA 94025 303-236-1292415-32P-5312

F. Singer (SAIC-D)R. Ra_ (USGS-D)

R, Spengler (USGS-D)B. Rd_son (LANL) MS 913, Box 25046MS D_3 Denver Federal CenterLos_os, NM 87545 Denver,CO 80225,FTS843-4318

E.Springer(LANL)G.RoBboom (USGS.R) MS J521703-64-4422 LosA/amos,NM 87545

505-667-9836J'.Roau,lshein(USGS-R)414_onal Center B.Steinkampf(USGS.D)Reston.,VA703-641t_ 169 O.Stirewalt (NRC)

CNWRA-NRCJ. Rousseau (USGS-D) 1235JeffersonDavis, Suite 1102303-236-5183 Arlington,VA 22202-3283

703-979-9129C. Sav_rd (USGS-NV)Box 32:7 .I.Smcldess(USGS-D)Mercury, NVFTS 575-5830 N.Smthmann(USGS-D)

S.SchillingCUSGS-D) E.Taylor(USGS-D)MS 913Denver Federal Center F. Thamir (USGS-D)Denver, CO 80225 MS 421, Box 25046303-236-0914 DenverFederalCenter

Denver,CO 8022,5U.Schimschal(USGS-D)

" M.Thompson(Sandia)D.Schmidt(USGS-D) 505-844-7146303-236-1262

R. Thompson38

MS 913DenverPed_ CenterDenver,CO 80225

/

W, Thordanon(USGS-D)303-236-5192

C. _kmorton (USGS-D)303-236-1106

V.TldweU(SLnd_)505-845-8956

N.Truk (USG$-R)_$ 959-5719

S. Wheat¢_ (DRI)POBox60220Reno,bP/ 89512702-673-7393

M. Whi_eld(U$OS.D)

W. Wt.tson(USGS-D)303-922-9107

AbbreviationsUSO$-R 'USG$ RestonUSOS.D. USOS DenverUSGS-M -USG$ MenloParkUSGS-S -USGS SacramentoUSGS-LV - USOS La=VegasLANL -LosAlamosNadonaiLabLBL LawrenceBerkolyLabMIT -"Mass.InstituteofTechnologyP_. PacificNorthwestLabCSM- ColoradoSchoolofMines

' L.LNL- LawrenceLiven'noraNadLabNRC. Nuclear RegulatoryCommissionUSNWTRB- USNuclearWasteTechnical,ReviewBoardBurec.D-Bureauof ReclamationDenverDR.[-DessertR_ea_h InstituteSAIC- ScienceApplicationsIntemadonalCorp,USDOE-LV - US Dept. of Energy, La.sVegas

39