Fault-induced damage controlling the formation of Carlin deposits

Fault-induced damage controlling the formation ofCarlin-type ore deposits

Steven Micklethwaite*Research School of Earth Sciences, Australian National University, ACT 0200, Australia

ABSTRACT

It is shown that the first-order control on the distribution of gold mineralization inthe northern Carlin Trend was fluid migration through fault-related damage net-works, triggered by slip events on the Post-Genesis fault system. This fault system con-sists of two segments ~5–7 km in length, linked at a stepover across a largegranodiorite intrusion. In the footwall of the system a wall-damage zone consists ofsteep, small-displacement faults spatially associated with mineralization. Manywall-damage zone structures formed on pre-existing planes of weakness (e.g. dikemargins, folded bedding planes and pre-existing faults). In active fault systems, thestatic stress changes that occur around an earthquake can be calculated using StressTransfer Modelling (STM) and used to understand the distribution of aftershock frac-turing on damage zone structures—the critical parameter being positive changes inCoulomb failure stress. In this study, STM was applied to model hypothetical slipevents on the Post and Genesis fault segments. The distribution of stress changematches the distribution of gold mineralization, with broad, shallow scallops of min-eralization occurring in the footwalls of the Post and Genesis fault segments, whereasmineralization at the tips of the fault system, the stepover and the hangingwall isdeeper (~0.7–0.9 km, 0.3–0.9 km and >1 km respectively). Static stress change calcula-tions indicate the fault system had the ability to induce damage, enhance permeabilityand tap fluids as deep as 15 km.

Key Words: faulting, stress change, fluid flow, permeability, Carlin, gold

INTRODUCTION

Gold mineralization from the northern Carlin trend formedat shallow crustal levels (Cline et al., 2005) during a potentiallyshort period of time (Hickey et al., 2009), adjacent to an en eche-lon, normal-offset, segmented fault system (Figure 1). Many is-sues remain controversial, such as the source of the gold, thesource of the fluids, and the nature of the physical processes thatcontrolled the migration of fluids. A particular question iswhether fluid migration was primarily controlled by reaction(e.g. reaction-enhanced porosity), lithology (e.g. variation inprimary porosity), pre-existing fracture networks, or short-livedfracturing due to active faulting. With regards to this question,the timing of mineralization has been linked to the onset of ex-tension in the Great Basin, during the mid to late Eocene, by anumber of independent geochronological studies (Hofstra et al.,1998; Ressel et al., 2000; Tretbar et al., 2000; Arehart et al.,

2003; Chakurian et al., 2003). This suggests faulting and frac-turing may have been a critical component influencing shallow,hydrothermal fluid migration and the formation of these giantCarlin-type gold deposits.

An important concept with regards to faulting is that slipevents and fault growth result in the development a fault coreand a damage zone (Chester and Logan, 1986). Fault cores arethose zones of gouge, localized slip surfaces and comminutedrock material across which the majority of slip is accommo-dated. Deformed wall rocks, containing networks of smallerfaults and fractures adjacent to the master fault, represent thedamage zones. Damage zones exist over wide areas (1–2 km)around large faults (Cochran et al., 2009) but the intensity ofdamage around master faults is a maximum at fault tips and inthe linking zones between fault segments (McGrath andDavison, 1995; Kim et al., 2004). Damage zones are crucialcomponents of a fault system, profoundly influencing the bulksealing properties of faults (Odling et al., 2004), the elasticproperties of wall rocks (Cochran et al., 2009), the orientations*E-mail: [email protected]

221

and magnitudes of near-fault stresses (Faulkner et al., 2006;Healy, 2008), and the localization of fluid flow (Chester and Lo-gan, 1986; Chen et al., 2007).

This paper reports results from an examination of the faultsystem associated with Carlin-type mineralization in the north-ern Carlin trend, north-east Nevada (Figure 1). A combinationof field mapping and static stress change modeling (King et al.,1994; Micklethwaite and Cox, 2006) were used to better under-stand the nature of damage zone structures and their relation-ship with fault-slip enhanced permeability and the distributionofmineralization. It is shown that damage zone structures likelyhad a critical first-order control on fluid migration, and thatfluid migration was dependent on active fault processes.

GEOLOGICAL SETTING

The geology of northeast Nevada has a protracted history(nicely summarized in Cline et al., 2005). Two of the most im-portant aspects of the geology of Carlin-type deposits in Ne-vada, originated early in the geological history of the region.The first aspect was rifting and passive margin development(late Proterozoic through to Devonian times) that produced thehost rocks for Carlin-type mineralization in the northern Carlintrend. The second aspect was development of the Antler orog-eny (initiating in the late Devonian), which resulted in thrustingof siliciclastic rocks overDevonian carbonates and shales by theRoberts Mountain Thrust. The majority of mineralization in thenorthern Carlin trend is hosted in the Devonian carbonates andshales beneath the Roberts Mountain Thrust. The thrust formssuch a significant feature that its hangingwall and footwallrocks are referred to as “upper plate” and “lower plate” respec-tively (although confusingly, upper plate rocks originate fromlower in the sedimentary sequence). Throughout the northernCarlin trend, isoclinal, chevron and open folds can be found inboth upper plate and lower plate rocks, probably originatingfrom emplacement of the Roberts Mountain allocthon.

Subsequent to theAntler orogeny,multiple orogenic eventsand periods of sedimentation affected central to northeast Ne-vada. During the Jurassic, back-arc plutonism and lesser lam-prophyre diking occurred, forming many of the intrusive rocksnow exposed in the Carlin trend (Figure1). Apatitethermochronology shows the Carlin trend underwent signifi-cant exhumation in the late Cretaceous, cooling rocks to<60-70∞C (Hickey et al., 2003) suggesting theywere at shallowdepths of <1–2 km by the time of mineralization. Carlin-typemineralization in the Eocene was coeval with renewed normalfaulting, high K calc-alkaline magmatism, and paleovalley for-mation involving minor northwest to west-northwest directedextension (Cline et al., 2005; Henry, 2008). In the northernCarlin trend, faults with apparently small and large normal off-sets (meters to thousands of meters) crosscut upper plate andlower plate rocks, and the Roberts Mountain Thrust. From thispoint on, the scope of the structural analysis in this paper ismainly limited to the fault architecture that has a spatial and

temporal relationship with Carlin-type gold mineralization inthe northern Carlin trend, presumably active during the mid tolate Eocene.

GOLDFIELD-SCALE FAULT GEOMETRY,KINEMATICS AND MINERALIZATION

The distribution of mineralization in Carlin deposits isknown to have two controls; namely stratigraphy and structure(e.g. Bettles, 2002). Gold mineralization is broadly stratiformon the scale of tens to hundreds of metres (Figure 2), but alsosystematically associated with steep-dipping, normal offsetfaults on the goldfield scale. In particular, themajority ofminer-alization in the northern Carlin trend is spatially associated withthe Post-Genesis fault and with small-displacement damagezone faults in its footwall (Figure 2). To a lesser extent minerali-sation is associated with fold hinges or enhanced fracturing incontact metamorphic zones around the Jurassic intrusion at thecentre of the trend (Figure 1), which likely provided arheological contrast (Bettles, 2002).

The Post-Genesis, Tara and Dee faults describe a NNW- toNNE-striking set of en echelon, left-stepping, faults (Figure 1).The Post-Genesis fault is a segmented structure, where thenorthern Post fault segment links across a granodiorite sill to thesouthernGenesis fault segment (Figure 1). The linkage zone is astepover containing a network of small-displacement faults(Heitt et al., 2003). The fault geometries are typical for linkageof fault segments, which interacted across the stepover andwerekinematically coherent (Walsh et al., 2003; Kim et al., 2004).Tracing the Post fault segment to the north, finds that the struc-ture branches into multiple splays and becomes difficult to mapnorth of Ren pit (Figure 1), suggesting the segment is ~7 kmlong. TheGenesis fault segment is ~5 km long,with reliable evi-dence of the fault trace disappearing in the vicinity of the Uni-versal Gas pit (Figure 1). Both faults dip steeply at 70–80°. Fornormal fault lengths that do not exceed the dimensions of thebrittle crust (10–15 km), the aspect ratio of down-dip fault widthto length is typically assumed to vary from one to three (faultsvarying from penny-shaped to elliptical; e.g. Scholz, 2002;Walsh et al., 2003). Thus the maximum present day depth of thePost and Genesis fault segments will be 4.7–6.9 km. Given thatthe present day erosion level of the Post-Genesis fault systemwas ~1 km and that the fault likely did not breach the surface(Hickey et al., 2003, Cline et al., 2005), then at the time of min-eralization the Post-Genesis fault reached depths of 7–8 km.This interpretation is more complicated if the Post and Genesisfault segments link to form a single surface in three dimensionsat depth (Walsh et al., 2003).

At the scale of the fault core, the Post fault segment is reallya zone consisting of several strands of closely spaced faults(Figure 2). Critically, Paleozoic rocks have apparent normal off-sets of up to 1000 m (Figure 2), but where the Post fault cuts theJurassic intrusion, offsets are ~60 m, indicating unusually highdisplacement gradients along the fault. In contrast, considering

222 Steven Micklethwaite

Faulting and Carlin-type gold 223

Figure 1. Map of the northern Carlin trend, modified from Moore (2002), highlighting the major faults associated with mineralisation, plus the Roberts MountainThrust (RMT), fold hinges and intrusions in the area. Bold fault lines represent the first-order faults interpreted to be active during mineralization. The Post-Gene-sis fault is ~14 km long and segmented, terminating just north of Ren pit and in the vicinity of Universal pit to the south. Note, southeast from the tip of the Genesisfault is the Castle Reef fault.Mapping inUniversal and Lantern open pits suggested theGenesis fault andCastle Reef fault did not link and that the Castle Reef faultmay not be as extensive as previously interpreted. Inset map shows the location of the trend relative to Nevada, USA.

the Miocene sediments in isolation, the western strands havesmall normal dip-slip offsets inMiocene sediments and the east-ern strand localized up to 200 m normal dip-slip offset (Figure2). These observations could be interpreted to argue that thema-jority of displacement was accumulated on the Post fault seg-ment prior to intrusion of the Jurassic granodiorite sill and thatthe Eocene and Miocene involved reactivation of a pre-existingnormal fault. However, the measured Paleozoic offset indicatesan unrealistically large fault length-displacement ratio of 0.14,relative to typical ratios of 0.03 (Schlische et al., 1996). Addi-

tionally, the Post fault segment is parallel with antiformal foldhinges close to the footwall of the fault core (Figure 1).

The Post fault has been previously interpreted as an in-verted normal fault (e.g. Cline et al., 2005;Muntean et al., 2007)but inverted normal faults result in hangingwall antiforms, notfootwall antiforms. Furthermore, such an interpretation wouldimply a plausible but very complicated geological history ofnormal fault formation, inversion during orogenesis, reactiva-tion as a normal fault pre-Jurassic intrusion and development ofhighly unusual displacement-length characteristics, followed


Figure 2. (A.)West to east cross-section from northern end of Goldstrike pit, modified from plate 3 of Thompson et al. (2002). Mineralization is broadly stratiformand associated with the Post fault segment and footwall faults. (B.) West to east cross-section throughMeikle deposit. The Popovich and Rodeo Creek formationsthin beneath the RobertsMountain Thrust at a fold in the thrust. The Post-Genesis fault core comprisesmultiple strands located along the east limb of the fold.Min-eralization (black dotted line) is concentrated in and adjacent to the Post-Genesis fault core, as a tongue high in the footwall of the Post-Genesis fault, and in UpperPlate rocks of the Post-Genesis fault core and hangingwall (800 m to >1000 m below surface). (C.) Orientation and lineation data of the Post-Genesis fault.

by further extensional reactivation in the Eocene and Miocene.Inspection of Figure 2b shows a relationship between the loca-tion of the fault plane, plus folding and pinching out of thePopovich and Rodeo Creek Formations, beneath the RobertsMountain thrust. Thus an alternative model is proposed,whereby during the Eocene portions of the Post fault nucleatedand propagated from the limbs of a pre-existing fold. In this par-ticular case, the fold may have formed at a ramp on a back thrustof the Roberts Mountain allocthon (Figure 3). The geometricconsequence of a fault forming on a pre-existing fold,or foldedthrust ramp, would be very large apparent displacement, eventhough actual displacementsmay be only a few hundredmetres.The geological history of the Post fault may be much simplerthan that previously assumed.

Kinematics

The slip direction on any fault parallels themaximum shearstress resolved on the fault plane, which can relate to thefar-field stress state (Ramsay and Lisle, 2000). The Post-Gene-sis fault preserves dip-slip lineations with small to moderateoblique components (Figure 2c), consistent with west-north-west directed extension, but caution is required interpretingthese lineations because they may represent overprinting move-ment from extension during theMiocene. However, the orienta-tion of the far-field stress state can be estimated from the orien-tations of syn-Eocene extensional faults, veins and dykes at ~40

Ma, mapped elsewhere in northeast Nevada. Mapping ofEocene sediments and volcanics has revealed north to northweststriking syn-Eocene faults, with small-displacement normaldip-slip faults crosscutting and tilting Eocene sediments andtuffs from paleovalleys (Cline et al., 2005; Henry, 2008). In ad-dition, there are steep dipping, epithermal quartz veins from theTuscarora mining district, dated at ~39.2 Ma, with an averagenortherly strike (Henry et al., 1998). Therefore the regional ex-tension direction was east-west to west-northwest directed at~40 Ma. Because the strike of the Post-Genesis fault system isnorth-northwest, this fault systemwas slightlymisoriented rela-tive to the extension direction in the Eocene. Stress and strainare tensors, which mean that west-northwest extension couldgenerate normal dip-slip with oblique-slip movement on thePost-Genesis fault when the maximum principal stress issubvertical and controlled by the lithostatic overburden. Theamount of oblique-slip realised during independent slip eventsis dependent on the ratios of the magnitudes of maximumprinciple stresses to one another, at the instant of failure(Ramsay and Lisle, 2000).

More specifically, within the northern Carlin trend thestrike of a subvertical Eocene dike (K-dike, Figure 1) indicateswest-northwest directed extension during intrusion. In contrast,the strikes of Eocene dikes close to the Genesis fault segmentappear to be rotated relative to the K-dike (Figure 1). The appar-ent rotation of Eocene dikes could be due to a oblique-slip com-ponent of movement on the Post-Genesis fault, or the dikes may


Figure 3. Conceptual model of Post fault formation, on the fold limb of a previous thrust ramp. The important point ofthis model is not whether there is a thrust ramp present, but that pre-existing folding of some type explains why thePost fault has such apparently large normal offset.

have intruded in that orientation due to the stress rotations thattypically occur towards the tips of fault segments.

Distribution of Mineralization

There are several important features of the distribution ofmineralization around the Post-Genesis fault system worth not-ing. (1) Themajority of ore comes from the Popovich formationin lower plate rocks (Figure 2) but mineralization crosses for-mation boundaries, and is found in many other rock types, in-cluding upper plate rocks above the Roberts Mountains Thrust

(e.g. Ren pit in the north of the trend, see also Figure 2). (2)Min-eralization in the footwall of the Post-Genesis fault system de-scribes broad but relatively shallow scallops, extending over 2km away from the segments. (3) Mineralization exists at depth(1.5–1.8 km) in the hangingwall of the fault, hosted in upperplate rocks (Figure 2). (4) Within and adjacent to the core of thePost-Genesis fault system,mineralisation deepens and becomesparallel with the fault (0.3–0.6 km). (5) Mineralization deepensin the stepover between the Post and Genesis fault segments(0.3–0.9 km), and has also been found at depths 0.7–0.9 km atthe tip of the Post fault (not shown in Figure 1).


Figure 4. (A.) Southwest to northeast cross-section through Betze-Post pit, showing opposing dips of small normaldip-slip faults relative to the Post fault segment. Section was modified from one supplied by Barrick Gold Ltd. Dataresolution is not good enough to determine trace of offset of faults through Upper Plate rocks. (B.) Schematic sectionand photo from a traverse on bench 5120, west wall of Betze-Post pit, showing general attitude of bedding horizonsand faults, in the Rodeo Creek formation of Lower Plate rocks. Note, stratigraphy on the formation scale is onlymildly folded but internally tight folding, bedding plane lineations and low angle detachments occur, possibly con-trolled by rheological contrasts between individual beds. Themajority of foldsmay have formed during emplacementof the Roberts Mountain Thrust, though a thorough study is required. (C.) Fault orientation and lineation data fromdamage zone structures exposed in both Betze-Post and Genesis pits.

So overall, a pattern emerges of shallow footwall mineral-ization, deepening along and parallel with themaster fault coresand stepover, and existing at depth in the hangingwall of the sys-tem. These observations also indicate that the distribution ofmineralization is not due to upper plate rocks acting as anaquitard (e.g. Cline et al., 2005) but rather that faulting and frac-turing (present in both upper and lower plate rocks) were impor-tant pathways for ore-forming fluids to access suitable reactivelithologies.

DAMAGE ZONE FAULTS

Damage zones are classified into different geometric typesdepending on their location around the master fault (Kim et al.,2004).Tip-damage zones are developed at the tips of faults, link-ing-damage zones form where adjacent fault segments interactand link (e.g. at stepovers), and wall-damage zones representfaulting and fracturing in the near-field, wall rock region of afault. The near-field region of a fault is defined here as that re-gionwithin one fault segment length of a fault (i.e. within ~5 kmof the Post-Genesis fault). The majority of mineralization re-lated to the Post-Genesis fault system is associated with wall-and linking-damage zone structures (Figure 1, see also Bettles2002; Heitt et al., 2003). Many of the wall-damage zone struc-tures in the Betze-Post and Genesis pits are off-fault (i.e. notnecessarily connected to the master fault), in the near-field tothe master fault and only exposed in the footwall because thehanging wall is covered by a Miocene sedimentary basin.

Typical characteristics of the damage zone faults are theirsteep dips (>60°), with multiple strike orientations and dip di-rections opposing those of the Post-Genesis fault (Figure 4).The average strike is semiparallel with the Post-Genesis fault(Figure 4c). A number are localised on the margins of dikes andsome are associated with folds. Old reverse faults are also pres-ent, which appear to have been reactivated during mineraliza-

tion (e.g. West Bazza fault in Betze-Post pit). Damage zonefaults generally have small, normal dip-slip offsets, and a rangeof lineation orientations. The Peculiar fault and East-EastLonglac fault are two examples from Betze-Post pit (Figure 5).Both faults are located on themargins of dikes, with smooth, un-dulating, semi-continuous main slip surfaces. East-EastLonglac fault is steeply dipping (155/68W) with normaldip-slip offset of meters to possibly a few tens of meters. Goodgold grades were present in and adjacent to the fault at theEast-East Longlac fault exposure shown in Figure 5b. Damagecan occur on all scales, and may have included the formation ofdistributed microcracking in mineralized sediments (Figure5c). Fault rocks originating from brecciated dike material (Fig-ure 5d), are commonly poorly consolidated, clay-rich withgritty textures and contain some narrow (>1 cm), anastamosing,dark shears (possibly ultracataclasites). Weak fabrics are vari-ably present. On the basis of global fault displacement-lengthdata (Schlische et al., 1996), many of the damage zone faults areunder-displaced (c.f. Vetel et al., 2005)—that is to say the dam-age zone faults are interpreted to be continuous for several kilo-meters (e.g.Moore, 2002) but they accommodate offsets of onlymeters to tens of meters.

In summary the damage zone faults tend to display a rangeof unusual characteristics—low displacement-length ratios,steep dips, dip directions opposing the master fault, strikesoblique to the extension direction in the Eocene, multiplelineation directions and spatial location on themargins of dikes,fold limbs or pre-existing reverse faults. Taken together theseobservations indicate damage zone faulting was strongly con-trolled by the mechanics of pre-existing structure or weak sur-faces, which determined their orientations. Indeed, in somecases there may even be multiple small-displacement faultsaligned along strike on the flanks of the same pre-existing dyke.Thus, the wall-damage zone to the Post-Genesis fault systemrepresents an unusual type of damage zone, where near-field


Figure 5. Field photographs of wall-damage faults. (A.) The Peculiar fault localised along the margin of a subvertical pre-existing dike, people for scale. (B.) TheEast-East Langlac fault localized on the hangingwall margin of a dike, with intense local visible alteration in the wall rock marked by silicification and bleaching.Good gold grades are associated with this exposure (C.Weakly pers comm). (C.) Hand specimen hangingwall sample, taken 3 m away from the fault showing pri-mary sedimentary features such as fining up, and cut by a distributed network of microcracks (m). (D.) Hand specimen from the hangingwall of the East-EastLanglac fault. Brecciated and altered dike material is associated with sooty pyrite (py) at its contact with silicified wall rock. A late barite vein (ba) transects thiscontact and entrains small fragments of pyrite. Sooty pyrite alteration is parallel with the fault and also infills microcracks.

faulting was controlled by the reactivation of pre-existingmisoriented structures, such as dikes, fold limbs and old faults.

STATIC STRESS CHANGE MODELING

Changes in Coulomb failure stress (DsF) are used as aproxy for the triggering of earthquakes and the distribution ofaftershocks (e.g. King et al., 1994; Stein et al., 1997; Parsons etal., 2008). Whenever a fault slip event occurs, highly connectedporosity is generated in the fault core and damage is reactivatedor created in wall rocks, enhancing permeability (Sheldon andMicklethwaite, 2007). On this basis, static stress change calcu-lations (otherwise referred to as Stress Transfer Modeling orSTM)may also be applied to predict domains where permeabil-ity was enhanced around fault systems.

STM has been successfully applied to explain the distribu-tion of fault-vein hosted mineralization around regional-scaleshear zones inmesothermal lode-gold deposits from greenstonebelts (Cox&Ruming, 2004;Micklethwaite&Cox, 2004, 2006;Micklethwaite et al., 2009). Three dimensional (3D) elastichalf-space dislocation models are used to calculate the stresschanges around faults. Critical parameters are (1) master faultgeometry, (2) kinematics (generally controlled by the orienta-tion of the far-field stresses), (3) depth of the master fault rela-tive to the Earth’s surface, and (4) the coseismic slip distributionon themaster fault plane, especially where slip events had a ten-dency to arrest at fault stepovers or fault tips.

Based on the field observations, the Post-Genesis fault sys-temwas reconstructed in 3D, with the top of the fault system setat 500 m below the model surface, and the fault surfaces subdi-vided intomultiple rectangular panels (Figure 6). The panels al-lowed bends in the fault surface to be approximated and slip ta-pered across the Post and Genesis fault segments (Figure 6a).The averagemagnitude of slip applied to each fault segmentwasdetermined according to fault size, using the empirical relation-ships presented by Wells and Coppersmith (1994). For faults inshales and calcareous sediments with low values of elasticmoduli (e.g. shear modulus of 1.4 ¥ 105 bars), average fault slipper event on the Post Fault andGenesis Fault segmentswould beexpected to be ~0.65 m. A series of “control model” experi-ments were also run, where the general stress change patternswere investigated around a single, planar, steeply dipping nor-mal dip-slip fault, with similar dimension, orientation and kine-matics to the Post fault segment. Comparisons of stress changesbetween the simpler and more complicated model configura-tions found that the first-order pattern of results was not greatlyaffected. Young’s modulus for shales is typically a low value of3 ¥ 105 bars. Poisson’s ratio is often taken to be 0.25 but for fluidsaturated rocks Poisson’s ratio may be as high as 0.4. Values forthe principal stresses s1>s2>s3 were fixed at 27.0 MPa/km,16.5 MPa/km and 10.0 MPa/km respectively, consistent withcrust undergoing extension that is in “frictional equilibrium”(Brudy et al., 1997). Either way, model results are generally in-sensitive to variations in elastic parameters and differential

stress (King et al., 1994), and experiments that varied these pa-rameters had no significant impact on results. The change inCoulomb failure stress was calculated using;

DsF = Dt + m¢(Dsn).

This simple equation states that Coulomb failure stresschange, adjacent to a fault slip event, is related to changes inshear stress (Dt), changes in normal stress (Dsn) and apparentfriction (m¢). Apparent friction is related to the coefficient ofstatic friction (Byerlee, 1978) and Skempton’s coefficient(Micklethwaite and Cox, 2006), and for shales or calcareousmuddy sediments the value used was m¢ = 0.14.

Results

Changes in Coulomb failure stress were calculated forfaults of specific orientations. Figure 6 shows the distribution ofpositive stress changes for (1) normal dip-slip surfaces whichare optimally oriented for failure relative to the far-field stressregime, and, (2) faults with an orientation parallel with the aver-age orientation of wall-damage zone structures (i.e. 160/75,rake -85, according to the right-hand rule). Those domainswhere positive stresses are resolved for optimally orientedfaults, represent volumes expected to contain the most intensedamage (i.e. in step-overs and at fault tips). Those domainswhere positive stresses are resolved for faults with typicalwall-damage zone orientations represent the volumes wherewall-damage zone type faults are most likely to have been acti-vated.

Figure 6 demonstrates that, for optimally oriented normaldip-slip faults, domains of positive stress change are concen-trated at the tips and in the stepover of the Post-Genesis faultsystem, forming dip-parallel, pipe-like volumes that are contin-uous to the base of the fault system. Thus, the presence of deepmineralization at the tip of the Post fault and in the Post-Genesisstepover could be explained by the triggering of optimally ori-ented normal faults, in domains of tip-zone damage. Positivestress changes at the Earth’s surface, deepen close to and alongthe Post fault segment. Figure 6b shows that, for faults with thesame orientation as wall-damage zone faults, shallow scallopsof positive stress change are generated in the footwall of thePost-Genesis fault system (matching the distribution of miner-alization in the Genesis and Betze-Post pits). Positive stresschanges also form semi-continuous domains that deepen closeto the fault surfaces (where mineralization is known to deepen),and inwell-developed domains in the hangingwall of the system(where mineralization has been found at depths of >1 km).These distributions of positive stress change bear marked simi-laritieswith the distribution of goldmineralization noted earlier,thus the mineralisation associated with the Post-Genesis faultsystem occurs in locations that are expected to have accumu-lated damage during slip events on that fault system. For exam-ple, the Genesis and Betze-Post pits both fit within the footwall



Figure 6. (A.) Post-Genesis fault model in 3D, composed of multiple panels with displacement distributed across the surfaces of the fault segments. (B.)Map viewof the faultmodel, shownwith the gradient ofCoulomb failure stress change used inC-D. (C.) 2D results at increasing depth slices, showing stress changes resolvedon optimally oriented normal dip-slip faults (010/60, rake -90, relative to the right-hand rule). Black polygons represent the surface projection ofmineralization as-sociated with the Post-Genesis fault system, grey polygons are mineralization associated with other faults in the northern Carlin trend. Mineralization deepensclose to the fault planes and in the stepover of the system. Deep mineralization has also been detected in the hanginwall and north tip of the Post fault segment (notshown). Stress changes indicate damage is generated continuously to depth at the tips and stepover of the system, and damage also accumulates close to and parallelwith the fault plane, to model depths of >1 km. (D.) 2D results at increasing depth slices, showing stress changes resolved on optimally oriented normal dip-slipfaults (160/75, rake -85, relative to the right-hand rule). Stress changes indicate wall-damage faults are triggered in shallow footwall scallops, and close to and par-allel with the fault plane.Wall-damage type faults are not triggered in the stepover domain or fault tip zones, but are likely to have been triggered in the hangingwallof the system. (E.) 3D representation of the results shown in C-D. The distribution of stress change is asymmetric between the Post and Genesis fault segments andreaches 15 km depths.

stress domains, the deeper Deep Star andDeep Post deposits arelocated at the Post-Genesis stepover, where stress changes arecontinuous with depth.

The pattern of stress change is asymmetric around thePost-Genesis fault system (Figure 6e), due to the larger size ofthe Post fault segment, and the obliquity that existed betweenthe orientation of the fault and the Eocene extension direction.3D isosurfaces of positive stress change show that positivestress changes could have been generated down to ~15 km (Fig-ure 6e). Due to the asymmetry of the system, the Post fault seg-ment was capable of generating positive stress changes overlarger volumes of rock relative to the Genesis fault segment.

DISCUSSION

Damage zones are critical components of a fault system(Chester and Logan, 1986; Odling et al., 2004; Cochran et al.,2009). In the Carlin-type ore deposits of the northern Carlintrend a combination of field observations and numerical model-ling indicate that wall-damage and linking-damage zone struc-tures controlled the migration of fluids and influenced the gold-field-scale location of mineralization. The damage is bothoff-fault and located in the near-field of the Post-Genesis faultsystem, and is unusual because it formed along pre-existingplanes of weakness, such as dike margins, fold limbs and oldreverse faults.

Damage zone faults associated with mineralization in thenorthern Carlin Trend, extend up to 2 km laterally from themas-ter fault, across strike. This width narrows significantly withdepth andmineralization becomes localized adjacent to the faultplane, and in stepover domains (e.g. Heitt et al., 2003). Further-more, deep mineralization has been discovered at the northerntip of the Post fault segment and in the hanging wall of the Postfault segment.

Stress Transfer Modelling can be used to explain the distri-bution of mineralisation in the Carlin trend. Positive changes instatic stress, from slip events on the Post-Genesis fault system,correlate with the distribution ofmineralization. This suggests amodel whereby static stress changes triggered slip on pre-exist-ing weak surfaces or faults, induced damage, and exerted afirst-order control on the enhancement of permeability. Someimportant implications of the modeling are that the depth extentof damage is potentially much greater than the dimensions ofthe fault system and could have reached to 15 km (i.e. a full faultsegment length below the base of successive ruptures). Faulttips and stepovers in a system are critical domains but there arealso important non-intuitive domains influenced by physicaland geometric factors such as fault shape and far-field stress re-gime. STM is a useful tool for understanding the non-intuitivedistribution of fault-related damage that controlled permeabil-ity enhancement and mineralization in any system(Micklethwaite and Cox, 2004, 2006). For example, in thePost-Genesis fault system, static stress changes provide a ratio-nale for understanding the footwall scallop of mineralization

represented by the world-class Betze-Post and Genesis pits.The approach outlined here provides a physically sound ba-

sis to understand fluidmigration and the distribution ofmineral-ization. On the meter scale physico-chemical processes such asreaction-enhanced porosity may well be important, which pos-sibly explains the presence of approximately stratiform miner-alization in places. However, the first-order control, on the gold-field scale, was faulting and fracturing related to slip events onthe Post-Genesis fault system. Fluid flow during Carlin-typemineralization has previously been interpreted as “passive”(Cline et al., 2005), meaning that mineralizing fluids exploitedpre-existing, quasi-permanent porosity networks, such as frac-tures and bedding horizons. However, the results of this studysuggest that porosity was generated by the faulting process, per-haps repeatedly.Mineralizationwas dependent on slip events onthe Post-Genesis fault system, which possibly resulted in cyclesof coseismic fracturing and interseismic healing(Micklethwaite et al., 2010). During and following a slip event,fluids migrated through the fault core and damage zone of thePost-Genesis fault system, before reacting with andmineralizing suitable lithological hosts.

ACKNOWLEDGEMENTS

ConnieNutt is thanked for a thoughtful review of themanu-script. Barrick Gold and Newmont, are thanked for excellent lo-gistical and financial support during the course of this study.Trevor Beardsmore, Francois Robert, Paul Doback, and LeeSampson of Barrick Gold, and Alan Goode of AMIRA Interna-tional were invaluable in helping the project start. Specialthanks go to CharlesWeakly and JimEssman for all their help inthe field.Most of thisworkwas completedwhile S.M.was at theAustralian National University working under ARC Linkagegrant LP0562164, with Prof. Stephen Cox.

REFERENCES

Arehart, G.B., Chakurian, A.M., Trebar, D.R., Christensen, J.N., McInnes,B.A. and Donelick, R.A., 2003, Evaluation of radioisotope dating ofCarlin-type deposits in the Great Basin, western North America, andimplications for deposit genesis: Economic Geology, v. 98, p.235–248.

Bettles, K., 2002, Exploration and Geology, 1962 to 2002, at the GoldstrikeProperty, Carlin Trend, Nevada. Economic Geology, Special Publica-tion, v. 9, p. 275–298.

Brudy, M., Zoback, M.D., Fuchs, K., Rummel, F. and Baumgartner, J., 1997,Estimation of the complete stress tensor to 8 kmdepth in theKTB scien-tific drill holes: Implications for crustal strength: Journal of Geophysi-cal Research, v. 102, p. 18,453-18,475.

Byerlee, J.D., 1978, Friction of rock: Pure and Applied Geophysics, v. 116, p.615-626.

Chakurian, A.M., Arehart, G.B., Donelick, R.A., Zhang, X. and Reiners, P.W.,2003, Timing constraints of gold mineralization along the Carlin trendutilizing apatite fission-track, 40Ar/39Ar, and apatite (U-Th)/He: Eco-nomic Geology, v. 98, p. 1159–1171.

Chen, W.D., Tanaka, H., Huang, H., Lu, C., Lee, C. and Wang, C., 2007, Fluid


infiltration associated with seismic faulting: Examining chemical andmineralogical compositions of fault rocks from the Chelungpu fault:Tectonophysics, v. 443, p. 243–254.

Chester, F.M. and Logan, J.M., 1986, Implications for mechanical properties ofbrittle faults from observations of the Punchbowl Fault Zone, Califor-nia: Pure and Applied Geophysics, v. 124, p. 79–106.

Cline, J.S., Hofstra, A.H., Muntean, J.L., Tosdal, R.M. and Hickey, K.A., 2005,Carlin-type gold deposits in Nevada: Critical geological characteristicsand viable models: Economic Geology 100th Anniversary Volume, p.451–484.

Cochran, E.S., Li, Y-.G., Shearer, P.M., Barbot, S., Fialko, Y. and Vidale, J.E.,2009, Seismic and geodetic evidence for extensive, long-lived faultdamage zones: Geology, v. 37, p. 315–318, doi:10.1130/G25306A.

Cox, S.F. andRuming,K., 2004, The St Ivesmesothermal gold system,WesternAustralia—a case of golden aftershocks? Journal of Structural Geology,v. 26, p. 1109–1125.

Faulkner, D.R., Mitchell, T.M., Healy, D. and Heap, M.J., 2006, Slip on ‘weak’faults by the rotation of regional stress in the fracture damage zone: Na-ture, v. 444, p. 922–925, doi:10.1038/nature05353.

Healy, D., 2008, Damage patterns, stress rotations and pore fluid pressures instrike-slip fault zones: Journal of Geophysical Research, v. 113,B12407, doi:10.1029/2008JB005655.

Heitt, D.G., Dunbar, W.W., Thompson, T.B. and Jackson, R.G., 2003, Geologyand geochemistry of the Deep Star gold deposit, Carlin Trend, Nevada:Economic Geology, v. 95, p. 1107–1135.

Henry, C.D., Boden, D.R., and Castor, S.B., 1998, Geology and mineralizationof the Eocene Tuscarora volcanic field, ElkoCounty, Nevada: U.S. Geo-logical Survey Open-File Report 98-338, p. 279–290.

Henry, C.D., 2008, Ash-flow tuffs and paleovalleys in northeastern Nevada:Implications for Eocene paleogeography and extension in the Sevierhinterland, northern Great Basin: Geosphere, v. 4, p. 1-35,doi:10.1130/GES00122.1.

Hickey, K.A., Haynes, S., Tosdal, R.M. andMortensen, J.K., 2003, Cretaceous-Paleogene denudation, volcanism and faulting in the Carlin-Jerritt Can-yon mining district, northeastern Nevada: Implications for thepaleogeographic and tectonic environment ofCarlin-type gold deposits,in Eliopoulos, D.G. et al., eds., Mineral exploration and sustainable de-velopment: Rotterdam, Millpress, p. 685–688.

Hickey, K., Dipple, G., Barker, S. andDonelick, R., 2009, In the blink of an eye:thermal contraints on the duration of hydrothermal fluid flow duringformation of the Carlin Au-deposits, USA, in Proceedings of the 10thbiennial SGA meeting of the society for geology applied to mineral de-posits, Townsville, Australia, p.282–284.

Hofstra, A.H., Rye, R.O., 1998, dDand d18Odata fromCarlin-type gold depos-its-Implications for genetic models: U.S. Geological Survey Open-FileReport 98-338, p. 202–210.

Kim, Y-.S. Peacock, D.C.P. and Sanderson, D.J., 2004, Fault damage zones:Journal of Structural Geology, v. 26, p. 503–517.

King, G.C.P., Stein, R.S. and Lin, J., 1994, Static stress changes and the trigger-ing of earthquakes. Bulletin of the Seismological Society of America, v.84, p. 935-953.

McGrath, A.G., and Davison, I., 1995, Damage zone geometry around faulttips: Journal of Structural Geology, v. 17, p. 1011–1024.

Micklethwaite, S. and Cox, S.F., 2004, Fault-segment rupture, aftershock-zonefluid flow, and mineralization: Geology, v. 32, 813–816.

Micklethwaite, S. and Cox, S.F., 2006, Progressive fault triggering and fluidflow in aftershock domains: Examples frommineralizedArchaean faultsystems: Earth and Planetary Science Letters, v. 250, p. 318–330.

Micklethwaite, S., Sheldon, H.A. and Baker, T., 2010, Active fault and shearprocesses and their implications for mineral deposit formation and dis-covery: Journal of Structural Geology, v. 32, p. 151–165.

Moore, S., editor, 2002, Geology of the north Carlin trend: Nevada Bureau ofMines and Geology, Bulletin 111, scale 1:24 000, 1 sheet.

Muntean, J.L., Coward,M.P. and Tarnocai, C.A., 2007, Reactivated Palaeozoicnormal faults: Controls on the formation of Carlin-type gold deposits innorth-central Nevada, in Ries, A.C. et al., eds., Deformation of the con-tinental crust: The legacy of Mike Coward, Geological Society of Lon-don Special Publication, v. 272, p. 573–589.

Odling, N.E., Harris, S.D. and Knipe, R.J., 2004, Permeability scaling proper-ties of fault damage zones in siliciclastic rocks: Journal of StructuralGeology, v. 26, p. 1727–1747.

Parsons, T., Ji, C., andKirby, E., 2008, Stress changes from the 2008Wenchuanearthquake and increased hazard in the Sichuan basin: Nature, v. 454,p.509–510.

Ramsay, J.G. and Lisle, R.J., 2000, The techniques of modern structural geol-ogy,Volume 3:Applications of continuummechanics in structural geol-ogy: Academic Press, 1061 p.

Ressel, M.W., Noble, D.C., Henry, C.D. and Trudel, W.S., 2000, Dike-hostedores of the Beast deposit and the importance of Eocene magmatism ingold mineralization of the Carlin Trend, Nevada: Economic Geology, v.95, p. 1417–1444.

Schlische, R.W., Young, S.S., Ackermann, R.V., and Gupta, A., 1996, Geome-try and scaling relations of a population of very small rift-related normalfaults: Geology, v. 24, p. 683–686.

Sheldon, H.A. and Micklethwaite, S., 2007, Damage and permeability aroundfaults: Implications for mineralization. Geology, v. 34, p. 903-906, doi:10.1130/G23860A.1.

Stein, R.S., Barka, A.A. and Dieterich, J.H., 1997, Progressive failure on theNorth Anatolian fault since 1939 by earthquake stress triggering, Geo-physical Journal International, v. 128, p. 594–604.

Thompson, T.B., Teal, L. and Meeuwig, R.O., editors, 2002, Gold deposits ofthe Carlin trend: Nevada Bureau of Mines and Geology, Bulletin 111,204 p.

Tretbar, D.R., Arehart, G.B. and Christensen, J.N., 2000, Dating gold deposi-tion in a Carlin-type gold deposit using Rb/Sr methods on the mineralgalkhaite: Geology, v. 28, p. 947–950.

Vetel, W., Le Gall, B. and Walsh, J.J., 2005, Geometry and growth of an innerrift fault pattern: the Kino Sogo Fault Belt, Turckana Rift (NorthKenya): Journal of Structural Geology, v. 27, p. 2204–2222.

Walsh, J.J., Bailey, W.R., Nicol, A. and Bonson, C.G., 2003, Formation of seg-mented normal faults: a 3-D perspective: Journal of Structural Geology,v. 25, p. 1251–1262.

Wells, D.L. and Coppersmith, K.J., 1994, New empirical relationships amongmagnitude, rupture length, rupture width, rupture area, and surface dis-placement: Bulletin of the Seismological Society of America, v. 84, p.974–1002.


Documents

Fault-induced damage controlling the formation of Carlin deposits