Active Normal Faulting Beneath a Salt Layer: An ...schlisch/aapg_ff_2000.pdfdevelopment of extensional forced folds above ... extensional forced folding and within the underly-

ABSTRACT

Scaled experimental models show that the pres-ence of a viscous layer, such as salt, facilitates thedevelopment of extensional forced folds aboveactive normal faults. The geometries of the exten-sional forced folds and their associated secondaryfault patterns depend on the thickness and viscosityof the viscous layer, the thickness of the coversequence, the strength and ductility of the coversequence, and the magnitude and rate of displace-ment on the underlying master normal fault.Increasing the thickness of the viscous layer and thecohesive strength and ductility of the overburdenenhances the decoupling between the deep andshallow deformation. Alternatively, increasing theviscosity of the viscous layer, the thickness of theoverburden, and the magnitude and rate of displace-ment on the master normal fault reduces the decou-pling between the deep and shallow deformation.Enhanced decoupling facilitates the formation ofbroad extensional forced folds and the developmentof detached secondary faults both near and far fromthe master normal faults. The model-predicteddeformation patterns closely resemble those observedin the Gulf of Suez, the Haltenbanken area of offshoreNorway, and the Jeanne d’Arc Basin of the GrandBanks, offshore southeastern Canada.

INTRODUCTION

Extensional forced folds commonly form abovenormal faults (Figure 1). Examples include flexures

627AAPG Bulletin, V. 84, No. 5 (May 2000), P. 627–651.

©Copyright 2000. The American Association of Petroleum Geologists. Allrights reserved.

1Manuscript received January 4, 1999; revised manuscript receivedAugust 25, 1999; final acceptance November 15, 1999.

2Mobil Technology Company, Dallas, Texas. Present address:Department of Geological Sciences, Rutgers University, Piscataway, NewJersey 08854; e-mail: [email protected]

3Mobil Producing Nigeria, Lagos, Nigeria.We thank Rolf Ackermann, Fred Bowes, Selena Dixon, Gloria Eisenstadt,

Karen Fredricks, Malin Ingebrigtsen, Roy Schlische, Ian Sinclair, Dan Shultz-Ela, Bruno Vendeville, and an anonymous reviewer for their help during thisstudy. In addition, Gloria Eisenstadt and Bruno Vendeville provided us withmany valuable ideas and insights on experimental modeling. Finally, wethank Mobil Technology Company for its support during this study.

Active Normal Faulting Beneath a Salt Layer: An ExperimentalStudy of Deformation Patterns in the Cover Sequence1

Martha Oliver Withjack2 and Sybil Callaway3

along the eastern and western margins of the Rhinegraben (Laubscher, 1982; Maurin, 1995), within theSuez rift (Brown, 1980; Colleta et al., 1988), andthroughout the Haltenbanken area of offshoreNorway (Withjack et al., 1989, 1990). In these threeregions, the presence of subsurface evaporites facili-tated the development of extensional forced folds bypartially decoupling the shallow, folded strata fromthe deep, faulted strata and basement (Laubscher,1982; Withjack et al., 1989, 1990; Maurin, 1995).

Hydrocarbons are commonly trapped within theshallow secondary structures associated withextensional forced folding and within the underly-ing fault blocks (Figure 1). Precisely locating thesestructural traps and identifying the small-scalestructures within them, however, can be difficult.Typically, the evaporitic packages associated withmany extensional forced folds have densities orvelocities that differ significantly from those of theadjacent strata, obscuring or distorting the seismicimage. Also, the structural geometries of manyextensional forced folds vary considerably withdepth, making seismic interpretation difficult. Abetter understanding of the deformation patternsassociated with extensional forced folding can pro-vide geoscientists with guidelines for interpretingfield, well, and seismic data. Specifically, this infor-mation can aid exploration and production effortsin extensional provinces with salt, including theNorth Sea, offshore Norway, the Gulf of Suez, off-shore eastern Canada, and the Gulf of Mexico.

We have used scaled experimental models to study the deformation patterns in the coversequence above an active master normal fault. Inour models, silicone polymer represented a saltlayer that deformed primarily by viscous flow. Drysand or wet clay simulated the sedimentary coverthat deformed primarily by cataclastic faulting.Previous workers also have studied extensionalforced folding using scaled physical models. Single-layer models composed of either sand or clay (e.g.,Vendeville, 1987; Withjack et al., 1990) and two-layer, sand/putty models (Vendeville, 1987) showedthat the dip of the master normal fault influencedthe deformation patterns associated with extension-al forced folding. Multilayer, sand/putty models

(e.g., Koyi et al., 1993; Nalpas and Brun, 1993;Vendeville et al., 1995; Higgins and Harris, 1997)suggested that other factors (e.g., the thickness ofthe viscous layer and overburden, the deformationrate, and the orientation of the master normal faultrelative to the extension direction) also affected thedeformation patterns associated with extensionalforced folding. We have supplemented this work bysystematically investigating how five variables (i.e.,the thickness of the viscous layer, the thickness ofthe cover sequence, the strength/ductility of thecover sequence, the magnitude of fault displace-ment, and the rate of fault displacement) inf lu-enced the deformation patterns associated withextensional forced folding. Our sensitivity analysisincluded a standard model and 11 alternate models,each with a single-parameter variation (Table 1).

The following sections describe the experimentalmaterials and procedure; review our modelingresults and compare them with those from previ-ous studies; and discuss how the models comparewith actual examples of extensional forced foldsfrom the Gulf of Suez, offshore Norway, and theGrand Banks of eastern Canada.

EXPERIMENTAL APPROACH

Modeling Materials and Procedure

The modeling apparatus had a horizontal metalbase, 33 cm wide and 61 cm long (Figure 2a). Alayer of homogeneous silicone polymer, represent-ing an evaporitic package, covered the entire metalbase. The silicone polymer (Rhodorsil gomme) wasa bouncing putty with a density of about 1150kg/m3 and a viscosity of about 104 Pa s (Weijermars,1986; S. Dixon, 1996, personal communication).

An upper layer of either homogeneous dry sandor very wet clay, representing the sedimentaryoverburden, covered the putty layer. Colored (butmechanically identical) sublayers within the upperlayer of sand or clay were used as marker horizonsto identify faults and folds. The dry sand in ourmodels was composed of quartz grains with diame-ters of less than 0.5 mm. Its density was about1600 kg/m3. It had a negligible cohesion and acoefficient of internal friction of about 0.6 (e.g.,Richard and Krantz, 1991; Nalpas and Brun, 1993;Vendeville et al., 1995). Localized cataclastic fault-ing (e.g., Rutter, 1986) was the primary deforma-tion style in the dry sand, even when strains weresmall. Thus, the dry sand had a negligible ductility

628 Faulting Beneath a Salt Layer

Table 1. Description of Models

Modeling Initial InitialMaterial and Putty Putty Overburden Displacement

Model Cohesion Viscosity Displacement Thickness Thickness Rate in cm/sName (Pa) (Pa s × 104) (cm) (cm) (cm) (cm/hr)

Standard Dry sand, ∼ 0 ∼ 3 1.41 1 3 8.18 × 10–6 (0.03)2 Dry sand, ∼ 0 ∼ 3 2.82 1 3 8.18 × 10–6 (0.03)3 Dry sand, ∼ 0 – 1.41 0 4 6.94 × 10–5 (0.25)4 Dry sand, ∼ 0 ∼ 3 1.41 2 3 8.18 × 10–6 (0.03)5 Dry sand, ∼ 0 ∼ 3 1.41 1 1 8.18 × 10–6 (0.03)6 Dry sand, ∼ 0 ∼ 3 1.41 1 1* 8.18 × 10–6 (0.03)7 Dry sand, ∼ 0 ∼ 1.5 5.66 1 3 8.18 × 10–6 (0.03)8 Dry sand, ∼ 0 ∼ 1.5 5.66 1 3 1.96 × 10–4 (0.71)9 Dry sand, ∼ 0 ∼ 1.5 5.66 1 3 3.93 × 10–4 (1.41)

10 Wet clay, ∼5 0 – 1.41 0 4 6.94 × 10–5 (0.25)11 Wet clay, ∼5 0 ∼ 1.5 1.41 1 3 6.94 × 10–5 (0.25)12 Wet clay, ∼5 0 ∼ 1.5 5.66 1 3 3.93 × 10–4 (1.41)

*Aggradation with overburden thickness increasing from 1 to 3 cm.

Figure 1—Sketch of an extensional forced fold above amaster normal fault. Dark gray layer represents anevaporitic package. Hydrocarbons are trapped withinthe secondary suprasalt structures produced by theforced folding and within the subsalt fault block.

as defined by Rutter (1986). The wet clay in ourmodels was composed predominantly of water(∼ 40% by weight) and kaolinite. Its density wasabout 1600 kg/m3. Recent analyses of the wet clay(Sims, 1993; Fugro-McClelland, Inc., 1992, personalcommunication; S. Dixon, 1996, personal communi-cation) showed that it had a low cohesion (∼ 50 Pa)and a coefficient of internal friction of about 0.5.Distributed cataclasis (e.g., Rutter, 1986) was the pri-mary deformation style in the wet clay when strainswere small. With increasing strain, however, defor-mation became localized and cataclastic faulting wasthe primary deformation style. Thus, the wet clayhad a greater ductility than that of the dry sand.

During the experiments, a constant downwardmovement on a precut surface, dipping 45°, withinthe metal base simulated displacement on a masternormal fault (Figure 2). In response, the putty layerflowed and the overlying layer of dry sand or wetclay faulted and folded. We filled in subsiding areaswith either sand or clay at regular intervals to simu-late deposition during deformation. These growthlayers initially had a f lat upper surface. Also, we

photographed the top surface of the models at reg-ular intervals to document surface deformationthrough time. After each experiment, we verticallysliced the model, creating serial cross sections. Thecross sections were photographed to record thefinal deformation state throughout the model.

Sensitivity Analysis

To determine the sensitivity of the modelingresults, we systematically varied the modelingparameters relative to a standard model (Table 1).In the standard model, the cover sequence consist-ed of dry sand, initially 3 cm thick. The initial thick-ness of the putty layer was 1 cm, the final displace-ment on the master normal fault was 1.41 cm, andthe displacement rate on the master normal faultwas 8.2 × 10-6 cm/s (0.03 cm/hr). In the alternatemodels, the initial thickness of the cover sequenceranged from 1 to 3 cm, the original thickness of theputty layer ranged from 0 to 2 cm, and the displace-ment on the master normal fault ranged from 0 to

Withjack and Callaway 629

Figure 2—Block diagramsof experimental apparatus(a) before and (b) afterfaulting. The black layer isputty, the white layer isdry sand or wet clayadded before faulting, andthe top gray layer in (b) issand or clay added duringfaulting.

4 cm

a

Master normal fault:45°-dipping precut

Cover sequence: homogeneous drysand or wet clay with marker horizons

Viscous layer:homogeneous

silicone polymer

Metal base

FootwallHanging wall

Growth layers ofhomogeneous sand or clay

b

33 cm

61 cm

5.66 cm. The displacement rate on the master nor-mal faults in the alternate models ranged from 8.2 ×10–6 to 3.9 × 10–4 cm/s (0.03–1.41 cm/hr). In sever-al of the alternate models, the cover sequence wascomposed of wet clay rather than dry sand.

Scaling

The cohesion and coefficient of internal frictionof the dry sand and wet clay were appropriate toensure dynamic similarity between the models andthe natural prototypes (see Appendix). The sandand clay, however, simulated different types ofrocks. The dry sand, with its negligible cohesionand ductility, represented rock that deformed pri-marily by localized cataclastic faulting. The wetclay, with its greater cohesion and ductility, repre-sented rock that deformed by both distributed cata-clasis and localized cataclastic faulting.

The displacement rate on the master normalfault in the models ranged from about 8.2 × 10–6 to3.9 × 10–4 cm/s (Table 1). These values for the dis-placement rate were selected to ensure dynamicsimilarity between the models and the natural pro-totypes (see Appendix). Models with low displace-ment rates simulated natural conditions where thevalue of P, the product of the fault displacementrate in nature and the viscosity of the evaporitic

layer, was low. Models with high displacementrates simulated natural conditions where the valueof P was high. Each model with its specified dis-placement rate represented a suite of natural condi-tions where P had the same value. For example,assuming that l* = 4 × 10–5 (i.e., 1 cm in the modelsrepresented 250 m in nature), our standard modelsimulated natural conditions where P = 2.2 × 108

Pa cm. Consequently, the standard model repre-sented natural conditions where the fault displace-ment rate was 2.2 × 10–10 cm/s and the evaporiticlayer had a viscosity of 1018 Pa s. It also representednatural conditions where the fault displacementrate was 2.2 × 10–9 cm/s and the evaporitic layerhad a viscosity of 1017 Pa s, or where the fault dis-placement rate was 2.2 × 10–8 cm/s and the evapo-ritic layer had a viscosity of 1016 Pa s.

RESULTS

Standard Model

Cross sections through the standard model showthat the putty layer thinned significantly near thefootwall cutoff and thickened slightly near thehanging-wall cutoff of the master normal fault(Figure 3a). A monocline, about 8 cm wide, formedin the cover sequence above the master normal


4 cm

4 cm b

Figure 3a

Edge offootwallblock

Edge ofhanging-wall

block

aFigure 3c

Sand added duringdeformationSand added beforedeformation

Silicone putty

Metal base c

Standard Model

Figure 3—Line drawingsand close-up photographof standard model: 1-cm-thick putty layer, 3-cm-thick sand overburden,1.41 cm displacement, anddisplacement rate of 8.2 × 10–6 cm/s. (a) Representative crosssection through standardmodel. (b) Map view ofstandard model.Dashed/dotted line showslocation of cross section.(c) Photograph of boxedarea of cross section in (a).

fault. The pregrowth marker beds were folded,dipping gently (<5°). The growth beds were alsofolded, thinning toward the footwall of the masternormal fault. A detached graben developed in thecover sequence near the footwall hinge of themonocline where bed curvature was greatest(Figure 3a, c). The detached graben was asymmet-ric with the main bounding fault dipping in thesame direction as the master normal fault. Thedetached graben spanned the length of the modeland had a relatively constant width along strike(Figure 3b). Although the detached graben andmaster normal fault were offset from each other,they were subparallel. This detached grabenaccommodated most of the lengthening associatedwith the horizontal component of displacement(i.e., heave) on the master normal fault. All of theabove modeling results were reproducible (i.e.,duplicate versions of the standard models had iden-tical structural geometries).

Displacement

Increasing the displacement on the master normalfault enhanced the existing structural geometries andcreated new structures (Figure 4). In model 2 with2.83 cm of displacement on the master normal fault

(Table 1), the folded beds dipped more steeply thandid those in the standard model. The detachedgraben was wider than the one in the standardmodel and was dissected by numerous minor nor-mal faults (Figure 4b). Also, the putty layer was sig-nificantly thinner near the footwall cutoff and thick-er near the hanging-wall cutoff of the master normalfault than in the standard model (Figure 4a, c). Adetached horst developed in the cover sequence inmodel 2 near the hanging-wall hinge of the mono-cline (Figure 4a, c).

Putty Thickness

Changing the putty thickness strongly affectedthe decoupling between the deep and shallowstructures (Figure 5). In model 3 without a puttylayer (Table 1), no monocline developed in thecover sequence. Instead, a graben formed directlyabove the master normal fault (Figure 5a, c). Thelarger graben-bounding fault linked directly withthe master normal fault. The smaller bounding faultemanated from the edge of the hanging-wall block.In model 4 with a 2-cm-thick putty layer (Table 1),deformation was widely distributed (Figure 5b, d).A broad monocline developed in the coversequence above the master normal fault. No


4 cm


Edge ofhanging-wall

block

c

a

b

Model 2


Silicone putty

Metal base

Figure 4c

Figure 4a

4 cm

Figure 4—Line drawingsand close-up photographof model 2. Model 2 hadthe same parameters as thestandard model except that displacement magnitude was 2.83 cm. (a) Representative crosssection through model 2.(b) Map view of model 2.Dashed/dotted line showslocation of cross section.(c) Photograph of boxedarea of cross section in (a).Note detached graben nearfootwall hinge and horstnear hanging-wall hinge ofmonocline. The smallfaults visible in the mapview were not identifiablein the cross sectional view.

detached graben formed near the footwall hinge ofthe monocline. Instead, a detached graben formednear the hanging-wall edge of the putty layer, morethan 16 cm from the master normal fault. This distal,detached graben accommodated most of the length-ening associated with the horizontal component ofdisplacement on the master normal fault. The loca-tion of this distal, detached graben was related to themodel boundary conditions (i.e., the edge of theputty layer). If model 4 (and its putty layer) had beenwider, then this detached graben may have formedeven farther from the master normal fault.

Overburden Thickness

Decreasing the overburden thickness accentuat-ed the decoupling between the deep and shallowstructures (Figure 6). In model 5 with a 1-cm-thick

overburden (Table 1), a narrow monocline devel-oped above the master normal fault (Figure 6a). Adetached graben formed locally near the footwallhinge of the monocline (Figure 6d). A second, longerdetached graben formed near the footwall edge ofthe putty layer, more than 16 cm from the masternormal fault. As discussed previously, the location ofthis detached graben was related to the modelboundary conditions (i.e., the edge of the puttylayer). If model 5 (and its putty layer) had beenwider, then this detached graben may have formedeven farther from the master normal fault. In model6, an aggradation model where the overburden thick-ness increased from 1 to 3 cm through time (Table1), deformation patterns more closely resembledthose of the standard model. A monocline developedabove the master normal fault, and detached normalfaults developed near the footwall hinge of the mon-ocline (Figure 6b, c). Locally, these detached normal


Edge ofhanging-

wall block

4 cm


4 cm

Edge ofhanging-

wall block


Model 3 Model 4

a

b

c d

Model 4

Model 3


Silicone putty

Metal base

Figure 6a

Figure 6b

4 cm

4 cm

Figure 5—Line drawings ofmodels 3 and 4. Models 3and 4 had same parametersas standard model exceptthat putty thicknesses were0 and 2 cm, respectively.(a, b) Representative crosssections through models 3 and 4, respectively. (c, d)Map views of models 3 and4, respectively. Dashed/dotted lines show locationsof cross sections.

faults produced a footwall graben similar to that inthe standard model (Figure 6c). The polarity of themain graben-bounding fault, however, changed alongstrike (Figure 6e). Near the center of model 6, themain detached fault dipped in the opposite directionas the master normal fault (Figure 6b). As in the stan-dard model, the detached normal faults in models 5and 6 accommodated most of the lengthening associ-ated with the horizontal component of displacementon the master normal fault.

Displacement Rate

We varied the displacement rate on the masternormal faults in models 7, 8, and 9. These models

also differed from the standard model in that the sil-icone putty had a slightly lower viscosity and thetotal fault displacement was significantly larger,5.66 cm (Table 1). As discussed previously and inthe Appendix, the models with low displacementrates simulated geologic conditions where fault dis-placement rates in nature were low or the evaporit-ic layer had a low viscosity (i.e., P was low). Themodels with high displacement rates simulatedgeologic conditions where fault displacement ratesin nature were high or the evaporitic layer had ahigh viscosity (i.e., P was high). Comparisons ofmodels 7, 8, and 9 show that the fault displacementrate in the models strongly affects the distributionof deformation within the cover sequence (Figures7c, 8, 9c, 10c).


4 cm

4 cm

4 cm

4 cm 4 cm

Edge ofhanging-

wall block


Edge ofhanging-

wall block


edModel 5 Model 6

c

a

b

Model 5

Model 6

Model 6


Silicone putty

Metal baseFigure 6a

Figure 6b

Figure 6c

Figure 6—Line drawingsof models 5 and 6. Models 5 and 6 had sameparameters as standardmodel except that over-burden thicknesses were 1 cm and aggrading from1 to 3 cm, respectively. (a) Representative crosssection through model 5.(b, c) Representative cross sections throughmodel 6. (d, e) Map viewsof models 5 and 6, respectively. Dashed/dotted lines show locations of cross sections.

In model 7 with the low displacement rate, defor-mation was broadly distributed. During the earlystages of the experiment, a very wide monoclinedeveloped above the master normal fault. Soonafter, a detached graben formed near the hanging-wall edge of the putty layer (Figure 7a). Initially, thisdistal, hanging-wall graben accommodated most ofthe lengthening associated with the horizontal com-ponent of displacement on the master normalfault. Eventually, it became inactive, and a newdetached graben developed near the footwall edgeof the putty layer (Figure 7b). Now, this distal foot-wall graben accommodated most of the lengthen-ing associated with the horizontal component of

displacement on the master normal fault. As dis-cussed previously, the location of these distaldetached grabens was related to the model bound-ary conditions. If model 7 (and its putty layer) hadbeen wider, then these distal detached grabens mayhave formed even farther from the master normalfault. Faults also developed in model 7 (Figure 7)directly above the master normal fault. This faultingincluded a normal fault (N1) that dipped in thesame direction as the master normal fault, and acurved fault (NR1) that had normal displacement atdepth and reverse displacement near the uppersurface of the model. These faults, together withthe monoclinal folding, accommodated the relief


a

4 cm

Edge offootwall block

Edge of hanging-wall block

b

c

4 cm

MODEL 7: 0.73-cm displacement YX

Y

X

N1

Y

X

NR2

N1

Y

X

NR1N1


NR2

N1

NR3


Sand added during deformation

Sand added before deformation

Silicone putty

Metal base

Figure 8a

Figure 7—Line drawings of cross sections (left) and map views (right) of model 7 through time: (a) restoration at0.73 cm displacement; (b) restoration at 3.27 cm displacement; and (c) final configuration at 5.66 cm displacement.Model 7 had a displacement rate of 8.2 × 10–6 cm/s. Cross section restorations in (a) and (b) are based on finalgeometries and surface photographs. See Appendix for restoration methodology. N denotes normal faults, and NRindicates curved faults with normal displacement at deep levels and reverse displacement at shallow levels. Num-bers show order of fault development. Black lines are active faults, whereas thin gray lines are inactive faults.Dashed/dotted lines in map views show cross section location.

associated with the vertical component of displace-ment on the master normal fault. With additionaldisplacement on the master normal fault, a puttydiapir rose beneath the expanding footwall graben(Figure 7c). Both N1 and NR1 moved down the sur-face of the master normal fault. NR1 became inac-tive, and new curved faults (NR2 and NR3) replacedit (Figures 7c, 8a).

In model 8 (Figure 9) with the high displacementrate, most deformation occurred near the masternormal fault. Initially, a monocline developed abovethe master normal fault (Figure 9a). Soon after, adetached graben formed near the footwall hinge ofthe monocline. The main graben-bounding fault(N1) was stationary relative to the footwall block ofthe master normal fault. The other graben-boundingfault (N2) moved laterally as the detached grabenwidened. Eventually, N2 became inactive, and a sim-ilar normal fault (N3) replaced it (Figure 9b). As theexperiment progressed, curved faults (NR1 andNR2), similar to those in model 7, formed above themaster normal fault. The detached graben contin-ued to widen, causing N3 to move laterally and NR1and NR2 to move down the surface of the masternormal fault. Eventually, they became inactive, andnew faults (N4, NR3, and NR4) replaced them(Figures 9c, 8b).

In model 9 (Figure 10) with the very high displace-ment rate, most deformation occurred directly abovethe master normal fault. Unlike models 7 and 8, littlemonoclinal folding occurred during the early stagesof model 9. Instead, a graben formed directly abovethe master normal fault (Figure 10a). The maingraben-bounding fault (N1) linked with the masternormal fault. The other graben-bounding fault (N2)

moved down the surface of the master normal fault.Eventually, N2 became inactive, and a similar normalfault (N3) replaced it (Figure 10b). This processrepeated itself during the model, creating a suite oftransient antithetic normal faults in the hanging wallof the master normal fault (Figures 10c, 8c).

Cohesive Strength and Ductility ofOverburden

In models 10, 11, and 12, the overburden was com-posed of wet clay rather than dry sand (Table 1). Asdiscussed previously, the wet clay had a greater cohe-sion and ductility than the dry sand and, thus, simulat-ed a more cohesive overburden that deformed byboth distributed cataclasis and localized cataclasticfaulting. In model 10 with no putty layer, a narrowmonocline developed above the master normal fault(Figure 11a, c). Numerous, minor synthetic and anti-thetic normal faults propagated upward from the sur-face of the master normal fault into the monocline. Inmodel 11 with a 1-cm-thick putty layer, a broad mono-cline developed above the master normal fault (Figure11b). A few, very minor normal faults cut the coversequence near the hinges of the monocline.Apparently, undetectable, small-scale deformationwithin the clay layer accommodated most of thelengthening associated with the horizontal compo-nent of displacement on the master normal. Comparisonsof the deformation patterns in clay models 10 and 11show that the presence of a viscous layer significantlydecreases the coupling between the deep and shallowdeformation, even when the overburden has a greatercohesion and ductility.


a cbFigure 8—Close-up photographs of fault zones in sand models with varying displacement rates: (a) model 7, (b)model 8, and (c) model 9. Photograph locations are shown in Figures 7, 9, and 10, respectively.

The deformation patterns in clay models 10 and11 differed considerably from those of similar sandmodels. The modeling parameters for clay model10 and sand model 3 were identical, except foroverburden composition (Table 1). A monoclinedeveloped in the clay model, whereas no foldingoccurred in the sand model. Numerous syntheticand antithetic normal faults formed in the claymodel, whereas a simple, asymmetric graben devel-oped in the sand model. Deformation patterns inclay model 11 also differed from those of similarsand models (e.g., the standard model). A few,minor normal faults formed in the cover sequencein the clay/putty model, whereas major normal

faults developed in the cover sequence in all of thesand/putty models. These comparisons betweenthe clay and sand models suggest that the cohesivestrength and ductility of the overburden stronglyinfluence the style and distribution of deformationin the cover sequence above a master normal fault.

The modeling parameters for clay model 12 andsand model 9 were identical, except for overbur-den composition (Table 1). During the early stagesof model 12, a monocline formed above the masternormal fault. As the experiment progressed, thefolded beds steepened, and numerous blind normalfaults cut the attenuated monoclinal limb (Figure12a). A detached normal fault zone (N1) formed


Figure 9—Line drawings of cross sections (left) and map views (right) of model 8 through time: (a) restoration at1.28 cm displacement; (b) restoration at 3.94 cm displacement; and (c) final configuration at 5.66 cm displacement.Model 8 had a displacement rate of 2.0 × 10–4 cm/s. Cross section restorations in (a) and (b) are based on finalgeometries and surface photographs. See Appendix for restoration methodology. N denotes normal faults, and NRindicates curved faults with normal displacement at deep levels and reverse displacement at shallow levels. Num-bers show order of fault development. Black lines are active faults, whereas gray lines are inactive faults.Dashed/dotted line in map views shows cross section location.

4 cm

a


Edge of hanging-wall block

b

c

NR1

N4

NR3NR4

4 cm

N2

NR2

N1

N4N1 N3

N1


Y

X


Y

X


Y

X

N2

N1

NR1

NR2

N4

N1

N3

N4

NR3 NR4

N1

Figure 8b



Silicone putty

Metal base

above the footwall of the master normal fault (Figure12a). With increasing displacement on the masternormal fault, the folded beds in the monocline steep-ened. The older normal faults that cut the monoclinallimb became inactive, and new normal faults replacedthem (Figures 12b, c; 13b). Additional detached nor-mal fault zones developed above the footwall of themaster normal fault (e.g., N2, N4) (Figures 12b, c;13a). Deformation patterns in clay model 12 differedsignificantly from those in sand model 9 (Figure 10).Deformation was more distributed and folding wasmore important in the clay model than in the sandmodel. The coupling between the deep and shallowdeformation was much greater in the sand modelthan in the clay model.

DISCUSSION

Summary of Modeling Results

Our work shows that the presence of a viscouslayer, such as salt, facilitates the development of exten-sional forced folds above active normal faults. Thegeometries of the extensional forced folds and theirassociated fault patterns depend on the thickness andviscosity of the viscous layer, the thickness of thecover sequence, the cohesive strength and ductility ofthe cover sequence, and the magnitude and rate of dis-placement on the master normal fault (Figure 14).

Our work suggests that, with no viscous layer,most deformation in the cover sequence occurs


Figure 10—Line drawings of cross sections (left) and map views (right) of model through time: (a) restoration at1.65 cm displacement; (b) restoration at 2.73 cm displacement; and (c) final configuration at 5.66 cm displacement.Model 9 had a displacement rate of 3.9 × 10–4 cm/s. Cross section restorations in (a) and (b) are based on finalgeometries and surface photographs. See Appendix for restoration methodology. N denotes normal faults, and NRindicates curved faults with normal displacement at deep levels and reverse displacement at shallow levels. Num-bers show order of fault development. Black lines are active faults, whereas gray lines are inactive faults.Dashed/dotted line in map views shows cross section location.


Edge ofhanging-wall

block

a4 cm

b

c

4 cm

N6

N2

N3

N1

N1

N1


Y

X


Y

X


Y

X

N2

N1

N3

N1

N6

N1

Figure 8c



Silicone putty

Metal base

directly above the master normal fault. With a vis-cous layer, the secondary deformation in the coversequence is more widely distributed. Typically, amonocline forms above the master normal fault.The width of the monocline increases as (1) thethickness of the viscous layer increases and (2) thedisplacement rate of the master normal fault or theviscosity of the viscous layer decreases. Our workalso indicates that, with a viscous layer, secondarynormal faults commonly form in the cover sequenceboth near and far from the master normal fault.These secondary faults detach within the viscouslayer and accommodate much of the lengtheningassociated with the horizontal component of dis-placement on the master normal fault. The lateraloffset between the detached secondary normalfaults and the master normal fault increases as (1)the thickness of the viscous layer or the cohesivestrength/ductility of the cover sequence increasesand (2) the thickness of the cover sequence, thedisplacement rate of the master normal fault, or theviscosity of the viscous layer decreases. The cou-pling or linkage between the master normal faultand the secondary structures increases as the dis-placement on the master normal fault increases.Typically, a monocline forms in the cover sequenceduring the early stages of movement on the master

normal fault. As the displacement on the masternormal fault increases, the viscous layer thins nearthe edge of the footwall block, and the detached,secondary normal faults within the cover sequencebegin to link with the master normal fault throughthe attenuated viscous layer. Also, more secondaryfaults develop within the cover sequence as the dis-placement on the master normal fault increases.

Previous Work

Previous workers have also studied upward-propagating normal faults and extensional forcedfolds using experimental models. Horsfield (1977),Tsuneishi (1978), Vendeville (1987, 1988), andWithjack et al. (1990) conducted a series of single-layer models with dry sand or wet clay. Vendeville(1987, 1988) ran a series of two-layer models withdry sand and silicone polymer. In these single-layerand two-layer models, the authors varied the dip ofthe master normal fault from 45 to 90° and observeddeformation patterns through time. More recently,Vendeville et al. (1995) published the results ofthree additional two-layer sand/putty models. Thesemodels differed significantly from each other withvariations in the putty thickness, the overburden


a

b

c

Model 11

Model 10

Clay added duringdeformationClay added beforedeformation

Silicone putty

Metal base

Figure 11c

4 cm

4 cm

Figure 11—Line drawingsand photograph of model10 (clay overburden with0-cm-thick putty) andmodel 11 (clay overburdenwith 1-cm-thick puttylayer). (a, b) Representativecross sections throughmodels 10 and 11, respectively. (c) Photographof boxed area of cross section through model 11.

thickness, the final displacement on the master nor-mal fault, and the fault displacement rate. Nalpasand Brun (1993) and Vendeville et al. (1995) usedthree-layer models to investigate salt tectonics dur-ing basement-involved extension. These modelshad a lower sand layer, a middle layer of siliconepolymer, and an upper sand layer. During the mod-eling, localized horizontal extension at the base ofthe models created two or more high-angle normalfaults in the lower sand layer. In response, the over-lying putty layer flowed and the upper sand layerexperienced folding and faulting.

Direct comparisons of the results of these pub-lished models with those from our models are diffi-cult because of significant differences in the model-ing parameters and procedures. Generally, however,

the published deformation patterns are similar toour deformation patterns when the modelingparameters are similar. To facilitate these compar-isons, we calculated the values of two parametersfor our models and for several published models: (1)the product of the fault-displacement rate in themodels and the viscosity of the putty layer (i.e., drm• µm) and (2) the ratio of overburden thickness toputty layer thickness. The results of the differentmodeling studies were remarkably consistent whenthese parameters were similar (Figure 15). Generally,if the ratio of overburden thickness to putty-layerthickness was low and the value of drm • µm was low,then deformation in the cover sequence was broadlydistributed and decoupling was more complete. Avery wide monocline developed above the underlying


4 cm

N1MODEL 12: 3.57-cm displacement YX

4 cm


Edge ofhanging-wall

block

a

b

c

N1 N2 N3N4

Y

X

N1 N2MODEL 12: 4.66-cm displacement YX

Y

X


Y

X

N1

N1

N2

N1

N2

N3Figure 13a

Figure 13b

Clay added during deformation

Clay added before deformation

Silicone putty

Metal base

Figure 12—Line drawings of cross sections (left) and map views (right) of model 12 through time: (a) restoration at3.57 cm displacement; (b) restoration at 4.66 cm displacement; and (c) final configuration at 5.66 cm displacement.Model 12 had a displacement rate of 3.9 × 10–4 cm/s. Cross section restorations in (a) and (b) are based on finalgeometries and surface photographs. See Appendix for restoration methodology. N denotes normal faults. Numbersindicate order of development. Black lines are active faults, whereas gray lines are inactive faults. Dashed/dottedline in map views shows cross section location.

master normal fault, and detached normal faults anddiapirs formed both near and far from the master nor-mal fault. If the ratio of overburden thickness to putty-layer thickness was high and the value of drm • µm washigh, then deformation in the cover sequence wasfocused near the master normal fault and couplingwas more complete. A monocline developed aboveand detached normal faults formed near the underly-ing master normal fault. These conclusions, based onFigure 15, agree with observations by Koyi et al.(1993), Koyi and Petersen (1993a, 1993b), Nalpas andBrun (1993), and Vendeville et al. (1995).

Exploration and Production Implications

In many extensional provinces with salt (e.g., theNorth Sea, offshore Norway, offshore eastern Canada,the Gulf of Suez), hydrocarbons are trapped withinthe shallow, secondary structures associated withextensional forced folding and within the underlyingfault blocks (Figure 1). Our modeling results providethe following guidelines for identifying and definingthese suprasalt and subsalt structural traps.

(1) Movement on a subsalt normal fault generallyproduces a monocline (i.e., an extensional forced fold)within the cover sequence above the salt. The width ofthe monocline can vary considerably, depending onthe thickness of the cover sequence, the salt thicknessand viscosity, and the displacement rate on the subsaltnormal fault (Figure 14). Numerous small-scale faults(normal and reverse) may cut and significantly attenu-ate the limb of the monocline.

(2) Growth beds associated with extensionalforced folding dip and thicken away from the activesubsalt normal fault.

(3) The thickness of the salt layer changes near thesubsalt normal fault. It thins near the footwall cutoffand thickens near the hanging-wall cutoff.

(4) With increasing displacement on the subsaltnormal fault, secondary normal faults form within thecover sequence. These detached, suprasalt normalfaults accommodate most of the lengthening associat-ed with the horizontal component of displacement(i.e., the heave) of the subsalt normal fault.

(5) The suprasalt normal faults rarely connect withthe subsalt normal fault. They are more likely to linkwith the subsalt normal fault if the salt is very thinand highly viscous, the displacement on the subsaltnormal fault is large, or the displacement rate on thesubsalt normal fault is high.

(6) The width of the zone of suprasalt normalfaulting can vary considerably (Figure 14). Suprasaltnormal faults are more likely to develop several kilo-meters away from the subsalt normal fault if thecover sequence is thin, the salt is thick and has a lowviscosity, or the displacement rate on the subsalt nor-mal fault is low. All evaporitic packages, such as theputty layer in our models, have depositional or struc-tural edges. Suprasalt normal faults are more likely toform at these edges if the width of the zone ofsuprasalt normal faulting is large or the subsalt nor-mal fault is near the edges of the evaporitic package.

The modeling results also have implications forcross section balancing in extensional provinces withsalt. Cross sections through extensional forced foldsbalance if all of the deformation in the cover sequenceis included (Figure 16a, b). Unfortunately, it is notalways possible to include all of this deformation.Cross sections based on seismic profiles of limited lat-eral extent (e.g., typical three-dimensional seismiclines) might not include the distal deformation in thecover sequence. Also, the small-scale deformationwithin the cover sequence might be overlookedbecause it is below seismic resolution. Cross sectionsthat omit some of the deformation in the coversequence do not balance (Figure 16c and d). Thesecross sections, however, might appear to balanceusing salt evacuation, rather than extensional forcedfolding, as the deformation template (Figure 16e).Although extensional forced folds and salt-evacuationstructures have similar structural geometries, theyhave very different structural histories. With exten-sional forced folding, the subsalt normal faults formlate, after salt deposition. With salt-evacuation struc-tures, the subsalt normal faults form early, during saltdeposition.

APPLICATIONS

Our model-derived deformation patterns resem-ble those observed in the Gulf of Suez, the Halten-banken area of offshore Norway, and the Jeanned’Arc Basin of offshore Newfoundland, Canada.These similarities suggest that the modeling results


ab

Figure 13—Close-up photographs of model 12: (a) foot-wall fault zone and (b) faulted monocline. Photographlocations shown in Figure 12.

might apply to many extensional forced folds. Themodeling results, however, have limitations: (1) thedry sand, wet clay, and silicone polymer in the mod-els were homogeneous, (2) only a single master nor-mal fault existed in the models and it was planar, (3)only a single, constant-thickness viscous layer exist-ed in the models, and (4) no other processes (i.e.,fault-block rotation, footwall uplift, differential com-paction) influenced fold development. Thus, themodeling results must be used cautiously whenthey are applied to geologic examples where thecover sequence has pronounced preexisting inho-mogeneities; multiple master normal faults exist orthe master normal faults are listric; multiple salt lay-

ers exist or the salt layer has flowed prior to subsaltfaulting; or if tilting, footwall uplift, or compactionhave significantly influenced fold development.

Gulf of Suez

Continental rifting began in the Suez area duringthe latest Oligocene–earliest Miocene and culminat-ed during the latest Early Miocene (e.g., Robson,1971; Garfunkel and Bartov, 1977; Patton et al., 1994;Schütz, 1994). As rifting waned during the middleand late Miocene, evaporitic deposition occurredwithin the subsiding Suez rift basin. The Tortonian


Thickness of viscous layer

Thickness of overburden

Displacement

Cohesive strength or ductility of overburden

Displacement rate and/or viscosity of viscous layer

Thinner Thicker

ThinnerThicker

LessMore

Less More

Less coupling, distributed deformation

Sand Sand/putty Sand/putty

More coupling, focused deformation

Sand/putty

Less coupling, distributed deformation,less faulting, more folding

More coupling, focused deformation,more faulting, less folding

Sand/putty

Sand

Clay/putty

Clay

Sand/puttySand/putty

Less coupling, less faulting, less foldingMore coupling, more faulting, more folding

LowerHigher

Less coupling, distributed deformation,less faulting, more folding

More coupling, focused deformation,more faulting, less folding

Sand/putty Clay/putty

Sand/puttySand/puttySand/putty

Sand/putty

More coupling, focused deformation

Less coupling, distributed deformation

Figure 14—Summary ofmodeling results.

South Gharib Formation, the primary evaporiticpackage in the Suez area, consists of massivehalite interbedded with anhydrite and shale(Moustafa, 1975; Brown, 1980; Schütz, 1994). Thethickness of the South Gharib Formation variessignificantly, locally exceeding 2 km in salt diapirsin the central and southern Suez r ift basin(Moustafa, 1975; Schütz, 1994). The South GharibFormation is underlain and overlain by evaporiticpackages, the Belayim and Zeit formations, respec-tively, each composed of interbedded anhydrite,salt, and shale.

The extensional forced folds within the Suez rifthave a variety of deformational styles. Deformationwithin the Paleozoic, Mesozoic, and Cenozoic stratabelow the South Gharib Formation is localized eitherwithin fault zones or narrow extensional forced folds(Robson, 1971; Garfunkel and Bartov, 1977; Thiebaudand Robson, 1979, 1981; Brown, 1980; Patton, 1984;Coffield and Schamel, 1987, 1989; Withjack et al.,1990; Patton et al., 1994). Deformation within theMiocene and younger strata above the South GharibFormation is more distributed (Brown, 1980; Colletaet al., 1988). At some locations, this distributed defor-mation includes monoclines above the subsalt normalfaults and detached, suprasalt normal faults near thefootwall hinges of the monoclines (Figure 17a, b). Atother locations, the distributed deformation includesvery broad monoclines above subsalt normal faultsand distal, detached, suprasalt normal faults, tens ofkilometers from the subsalt normal faults (Figure 17c,

d). At these latter locations, it is likely that geologicconditions were favorable for decoupling and dis-tributed deformation (Figure 14). The South GharibFormation was thick and, with its predominantlyhalite composition, probably had a low viscosity. Theoverburden above the South Gharib Formation wasthin, and fault-displacement rates were low as riftingwaned during the late Miocene and Pliocene.

Haltenbanken Area, Offshore Norway

Several extensional events affected the Halten-banken area from the late Paleozoic to the earlyCenozoic (e.g., Bukovics and Ziegler, 1985; Blystad etal., 1995). Rifting occurred during the late EarlyPermian or Early to Middle Triassic. Widespread sub-sidence and deposition of evaporites, coals, and clas-tic sediments followed this early extensional episode.The Upper Triassic evaporitic package consists of atleast two layers of rock salt separated by clastic sedi-mentary rocks. Seismic and limited well data suggestthat locally the layers of rock salt and interveningclastic sedimentary rocks are each several hundredmeters thick (Jacobsen and van Veen, 1984; M.Ingebrigtsen, 1997, personal communication). TheUpper Triassic–Jurassic overburden above theevaporitic package is about 2–2.5 km thick. Riftingresumed during the late Middle Jurassic and contin-ued into the Early Cretaceous (Neocomian). Littledeposition occurred in the Haltenbanken area during


Standard, 2V

NB

5

0 2 4 6 8 100

10

20

Initial overburden thickness/putty-layer thickness

8

4

NBNB

NB

7NB

V

Less coupling;distributed deformation --

broad forced folding anddetached faulting and

diapirism near and far frommaster normal fault

More coupling;focused deformation --

forced folding and detachednormal faulting near master

normal fault

(Pa cm) d • µ

9

30

40

50

NB

rm m

fault-displacement rate •putty viscosity

Figure 15—Comparison ofpublished results withthose from sand modelsin our study. Vertical axis is the product offault-displacement rateand putty viscosity. Horizontal axis is theratio of overburden thickness to putty-layerthickness. Black andwhite circles indicate sandmodels with distributedand focused deformation,respectively. Numberedcircles are our models,circles with NB are models from Nalpas andBrun (1993), and circleswith V are models fromVendeville et al. (1995).

this Early Cretaceous extensional episode. Additionalrifting occurred during the Late Cretaceous and dur-ing the latest Cretaceous and Paleocene. The latterextensional episode directly preceded continentalbreakup in the North Atlantic Ocean.

Most extensional forced folding in the Halten-banken area occurred during the late Middle Jurassic–Early Cretaceous extensional episode. During thisevent, normal faults formed within the basement andPermian and Triassic strata below the salt. Simul-taneously, broad extensional forced folds developed inthe Upper Triassic–Jurassic strata above the salt.Apparently, the Upper Triassic evaporitic package par-tially decoupled the deep and shallow deformation(Withjack et al., 1989, 1990) (Figure 18). Deformationbelow the Upper Triassic evaporitic package is local-ized within fault zones, whereas deformation abovethe evaporitic package is distributed in zones that areup to 10 km wide. This distributed deformationincludes broad monoclines above subsalt normalfaults, and detached, suprasalt normal faults andgrabens near the footwall hinges of the monoclines

(e.g., the Mikkel structure, Figure 18b). The normalfaults that bound the grabens converge near the topof the Upper Triassic evaporitic package.

Jeanne d’Arc Basin, Grand Banks, OffshoreSoutheastern Canada

The Grand Banks area has had a complex Meso-zoic extensional history. Most workers agree that thefirst rifting episode began by the Late Triassic andcontinued into the Early Jurassic (e.g., Tankard andWelsink, 1987; Sinclair, 1993; Driscoll et al., 1995;Sinclair, 1995a, b). Based on seismic data from thesouthern Jeanne d’Arc Basin, we interpret that thisfirst rifting episode probably continued into theearly Middle Jurassic. Recent work (e.g., Sinclair,1993; Driscoll et al., 1995; Sinclair, 1995a, b) indi-cates that a second rifting episode began during thelate Late Jurassic and continued, perhaps intermit-tently, throughout the Early Cretaceous until conti-nental breakup.


aSalt

e

d

c

b

Salt

Figure 16—Restoration of extensional forced folds. (a) Cross section including all deformation associated withextensional forced folding. (b) Corresponding restoration. In this restoration, the lengths of the beds above andbelow the salt are identical (i.e., the cross section is balanced). (c) Cross section omitting the distal deformationassociated with extensional forced folding. (d) Corresponding restoration. In this restoration, the lengths of thebeds above the salt are greater than those below the salt (i.e., the cross section is not balanced). (e) Alternativerestoration with salt evacuation, rather than extensional forced folding, as deformation template. With the salt-evacuation template, it is assumed that a thick evaporitic package was deposited in the hanging wall of the normalfault and later evacuated. Cross sections that exclude the distal deformation associated with extensional forced fold-ing appear to balance with a salt-evacuation template.

Rift basins in the Grand Banks area, including theJeanne d’Arc Basin, contain thick packages of evap-oritic sedimentary rocks. Evaporitic depositionbegan during the Late Triassic and peaked duringthe early Early Jurassic with widespread depositionof halite (i.e., the Argo Formation) (Holser et al.,1988). The Upper Triassic and Lower Jurassic evap-oritic section consists of halite beds intercalatedwith shales, dolomites, and limestones. Massivehalite occurs in the upper part of the section. Abasalt layer (early Early Jurassic in age) commonlyexists within the evaporitic package. In the Jeanned’Arc Basin, the thickness of the Upper Triassic–Lower Jurassic evaporites locally exceeds 2 km(Holser et al., 1988). The Lower Jurassic evaporitesappear to occur throughout the Jeanne d’Arc Basin.

The Upper Triassic evaporites occur within thesouthern Jeanne d’Arc Basin, but it is unknownwhether they are present farther to the north.

The Cormorant structure is a broad anticline inthe southern Jeanne d’Arc Basin (Figure 19a). Thedeformation in the Lower–Middle Jurassic strataabove the Triassic–Jurassic evaporites includes (1)the Cormorant fold, (2) detached, suprasalt normalfaults near the footwall hinge of the fold, and (3)distal, detached, suprasalt normal faults near thesoutheastern edge of the salt basin. This suprasaltfolding and faulting developed from the late EarlyJurassic into the Middle Jurassic (Figure 19b, c).Seismic data clearly show that west-dipping nor-mal faults exist beneath the Cormorant structurebelow the Triassic–Jurassic evaporites. These


0

1

2

5 Km

SENW

seconds

SouthGharib Fm.

SENW

Zeit Fm.

5 Km

SENW

SENW

5 Km

d

c

a

b

Distal, detach, suprasaltnormal faults Monocline

Detached, suprasalt normalfaults & monoclines

Belayim Fm.

Deposited duringdeformation

Evaporite layer

Subsalt strata andbasement

No vertical exaggeration

Displayed 1:1 at 4 km/s

No vertical exaggeration

Figure 17c

5 KmNo vertical exaggeration

Figure 17b

Figure 17—Cross sections from Suez rift based on field, well, and seismic data. (a) Regional cross section across cen-tral Suez rift (after Colleta et al., 1988; Patton et al., 1994). (b) Line drawing of seismic section from Patton et al.(1994). Section location is shown in (a). (c) Cross section through Ramadan oil field (after Brown, 1980). Sectionlocation is shown in (d). (d) Regional cross section across southern Suez rift (after Colleta et al., 1988).

subsalt faults are antithetic to the Murre boundaryfault of the Jeanne d’Arc Basin. We propose thatthe Cormorant structure is, in part, an extensionalforced fold that formed above the subsalt, antithet-ic normal faults. Salt movement associated withdifferential sedimentary loading also might havecontributed to its development. It is likely that,during the Early–Middle Jurassic, geologic condi-tions favored the decoupling of the deep and shal-low deformation. The Triassic–Jurassic evaporiticpackage was thick, and the cover sequence abovethe evaporitic package was thin. The subsalt, anti-thetic normal faults below the Cormorant structureprobably had low displacement rates compared tothe Murre fault, the main boundary fault of theJeanne d’Arc Basin.

The Flying Foam structure is a large, faulted anti-cline in the northern Jeanne d’Arc Basin (Figure20a). The anticline affects the Jurassic and LowerCretaceous strata above the Triassic–Jurassic evapor-ites. A major, basement-involved, east-dipping nor-mal fault deforms the strata and basement below theTriassic–Jurassic evaporites. Enachescu (1987) inter-preted the Flying Foam structure as an immature saltdiapir, whereas Tankard et al. (1989) explained thestructure as a fault-bend fold related to a ramp-flatgeometry on the bounding fault of the northern

Jeanne d’Arc Basin. Based on our modeling resultsand our study of the seismic data, we propose a thirdinterpretation: the Flying Foam structure is an exten-sional forced fold above a subsalt normal fault.During Triassic–Jurassic rifting, the Mercury faultwas the main boundary fault of the northern Jeanned’Arc Basin (Figure 20c). Deposition of the LateTriassic–Early Jurassic evaporites and the overlyingJurassic strata occurred to the east of this base-ment-involved normal fault. As rifting intensifiedduring the Early Cretaceous, an east-dipping, base-ment-involved normal fault (possibly the northerncontinuation of the Murre fault) developed just tothe east of the Mercury fault (Figure 20b). TheTriassic–Jurassic evaporites decoupled the shallowand deep deformation. In response, an extensionalforced fold (i.e., the Flying Foam structure) formedin the Jurassic and Cretaceous strata above the sub-salt normal fault. Also, a major, detached, suprasaltnormal fault developed near the western boundaryof the Jeanne d’Arc Basin at the depositional edge ofthe Triassic–Jurassic evaporitic package. As in ourmodels, this detached, suprasalt normal fault devel-oped at the edge of the viscous layer and accommo-dated most of the lengthening associated with thehorizontal component of displacement on the sub-salt normal fault.


Figure 18—Haltenbankenarea, offshore Norway. (a) Line drawing of time-migrated seismic section (NRGS84-407)across Smørbukk structure,offshore Norway (afterWithjack et al., 1989, 1990). (b) Line drawing of time-migrated seismicsection (GMNR94-310)across Mikkel structure,offshore Norway.

CONCLUSIONS

We have used scaled experimental models tosystematically study the influence of five variables

on the deformation patterns associated withextensional forced folding. Our work, togetherwith the results of previous studies, shows thatthe presence of a viscous layer, such as salt, facilitates


Figure 19—Southern Jeanne d’Arc Basin, Grand Banks, offshore Newfoundland. (a) Line drawing of time-migratedseismic section 85-4A. (b) Restoration for the Middle Jurassic during deposition of the Whale member. SeeAppendix for restoration approach. Cormorant fold developed above subsalt normal faults, antithetic to Murreboundary fault. Some detached, suprasalt normal faults formed far from the subsalt normal faults. (c) Restorationfor the Early Jurassic during deposition of upper evaporitic package (i.e., Argo Formation).

5 Km

Argo Fm. -- beddedLower Jurassic

evaporites

Lower Jurassic basalt

0

2

4

6

8

NW SE

Cenomanian Unconformity

?

Upper Triassic strata (possiblyincluding evaporites)

?

Lower to MiddleJurassic strata

Murre G-87 85-4A Cormorant N-83

seconds

c

b

a

Strata deposited during deformation

Evaporitic package

Subsalt strata and basement

Middle to UpperTriassic strata


Monocline; growth beds thinningtoward footwall of subsalt norma faults Distal, detached, suprasalt normal faults

Strata deposited after deformation


HBV83-1950

2

4

6

8seconds

W E

5 Km

Tithonian unconformity

Aptian unconformity

Albianunconformity

?

water

a

b

cMercury fault

Upper Triassic/Lower Jurassicevaporites

Strata depositedduring deformationOverburden beforedeformation

Evaporitic package

Subsalt strata andbasement


Strata depositedafer deformation

Mercury fault

Mercury fault

Northernpropagation ofMurre fault (?)

Attenuated fold limb cut bysecondary normal faults

Detached normal fault nearwestern limit of salt deposition

Western limitof salt

deposition

Figure 20—Northern Jeanne d’Arc Basin, Grand Banks, offshore Newfoundland. (a) Line drawing of time-migrated seis-mic section HBV83-195 from northern Jeanne d’Arc Basin. (b) Restoration for the Aptian. See Appendix for restorationapproach. The Flying Foam structure formed above an east-dipping normal fault (possibly the northern continuation ofthe Murre fault). A major detached normal fault formed near the Mercury fault at the western limit of the Triassic–Jurassicevaporite package. (c) Restoration for the Late Jurassic. The Mercury fault was the boundary fault of the northernJeanne d’Arc prior to the Early Cretaceous. Triassic–Jurassic evaporites filled the basin during this early episode of rift-ing, and several kilometers of Lower–Upper Jurassic strata covered the evaporitic package.

the development of extensional forced folds. Ourwork suggests that the geometry of the extension-al forced fold and the secondary fault patterns thatdevelop during folding strongly depend on the fol-lowing variables:

(1) Salt thickness. With no salt layer, deforma-tion in the cover sequence occurs directly abovethe master normal fault, and most secondary nor-mal faults link with the master normal fault. With asalt layer, deformation in the cover sequence ismore widely distributed. Typically, a broad mono-cline forms above the master normal fault, anddetached, secondary normal faults form within thecover sequence. The fold is wider, and the sec-ondary normal faults develop farther from the mas-ter normal fault with a thick salt layer than with athin salt layer.

(2) Overburden thickness. Deformation in thecover sequence is more widely distributed with athin overburden than with a thick overburden.The fold above the master normal fault is wider,and the secondary normal faults in the coversequence develop farther from the master normalfault.

(3) Cohesive strength and ductility of the over-burden. Deformation in the cover sequence ismore widely distributed with a more cohesive,more ductile overburden than with a less cohe-sive, less ductile overburden. The fold above themaster normal fault is wider, and the secondarynormal faults in the cover sequence develop far-ther from the master normal fault.

(4) Fault displacement. As the displacement onthe master normal fault increases, the salt layerthins near the edge of the footwall block, anddetached, secondary normal faults within thecover sequence are more likely to link with themaster normal fault through the attenuated saltlayer.

(5) Fault displacement rate or salt viscosity.Deformation in the cover sequence is more widelydistributed with low values of fault displacementrate or salt viscosity than with high values of faultdisplacement rate or salt viscosity. The monoclineabove the master normal fault is wider, and thesecondary normal faults develop farther from themaster normal fault.

Our model-derived deformation patterns close-ly resemble those observed in the Gulf of Suez,the Haltenbanken area of offshore Norway, andthe Jeanne d’Arc Basin of the Grand Banks, off-shore southeastern Canada. These similarities sug-gest that the modeling results apply to manyextensional forced folds and can provide guide-lines for interpreting field, well, or seismic data inextensional provinces with salt.

APPENDIX

Scaling: Deformation by Cataclastic Faulting

The strength of most upper crustal rocks increases with depth,obeying a Mohr-Coulomb criterion of failure (e.g., Byerlee, 1978).According to this criterion,

(1)

where τ and σn are, respectively, the shear and normal stresses ona potential fault surface, c is the cohesion, and f is the coefficientof internal friction. This empirical criterion of failure describes theinitiation of new faults rather than the frictional reactivation ofexisting faults. For most sedimentary rocks, f ranges from about0.55 to 0.85 (e.g., Handin, 1966; Byerlee, 1978). For intact sedi-mentary rocks, c is about 10–20 MPa (Handin, 1966), whereas forhighly fractured sedimentary rocks, c is significantly less (e.g.,Byerlee, 1978; Brace and Kohlstedt, 1980). To ensure dynamicsimilarity between the models and natural prototypes, two condi-tions must be satisfied. First, the modeling materials and the rocksin nature must have similar coefficients of internal friction (e.g.,Nalpas and Brun, 1993; Weijermars et al., 1993). This condition issatisfied with either dry sand or wet clay as the modeling material.Second,

(2)

where c*, ρ*, g*, and l* are model-to-natural prototype ratios forcohesion, density, gravity, and length, respectively (e.g., Hubbert,1937; Weijermars et al., 1993; Vendeville et al., 1995). In our mod-els, the values of ρ* and g* were about 0.7 and 1, respectively, andl* was 10-4 to 10-5 (i.e., 1 cm in the models equaled 100–1000 m innature). Thus, to ensure dynamic similarity between the modelsand natural prototypes, the cohesion of rock must be about 104 to105 greater than that of the modeling materials. This condition wassatisfied with either dry sand or wet clay as the modeling material.The dry sand and wet clay simulated different types of rocks. Thedry sand, with its negligible cohesion and ductility, representedweak rock that deformed primarily by localized cataclastic fault-ing. The wet clay, with its slightly greater cohesion and ductility,represented slightly stronger rock that deformed by both distribut-ed cataclasis and localized cataclastic faulting.

Scaling: Deformation by Viscous Flow

To ensure dynamic similarity between the models and naturalprototypes when strata deform by viscous flow,

(3)

where dr* and µ* are model-to-natural prototype ratios for displace-ment rate and viscosity, respectively (e.g., Hubbert, 1937;Weijermars et al., 1993; Vendeville et al., 1995). Reformattingequation 3 yields,

(4)

where drm is the displacement rate on the master normal faults inthe models, µm is the viscosity of the putty layer in the models,and P = drn • µn, the product of the displacement rate on the mas-ter normal fault in nature and the viscosity of the salt layer.

We used equation 4 to determine appropriate values for thedisplacement rate on the master normal fault in our models. As

d g l Prm m= •( ) • ( ) •∗ ∗ ∗ρ µ2

d g lr∗ ∗ ∗ ∗ ∗= • • ( )

ρ µ2

c g l∗ ∗ ∗ ∗= • •ρ

τ σ= + •c f n


stated previously, the values of ρ*, g*, and l* were about 0.7, 1, and10-4 to 10-5, respectively. The value of µm was either 1.5 × 104 or3.0 × 104 Pa s (Table 1). The values of drn and µn and, consequent-ly, P can vary considerably. Studies from the Rhine and Baikal rifts(e.g., Kukal, 1989), the Newark basin of the eastern United States(Olsen et al., 1989, 1996), the Corsair fault of offshore Texas (e.g.,Barrientos, 1989), and several additional extensional basins (Nicolet al., 1997) suggest that long-term displacement rates on normalfaults in nature (drn) range from about 10-11 to 10-8 cm/s (i.e., 10-4

to 10-1 cm/yr). Under most geologic conditions, pure rocksaltbehaves as a Newtonian fluid whose viscosity depends on watercontent, grain size, and temperature (e.g., Urai et al., 1986; Nalpasand Brun, 1993; Weijermars et al., 1993; Vendeville et al., 1995).Depending on geologic conditions and material properties, the vis-cosity of a pure rocksalt layer (µn) ranges from less than 1016 tomore than 1019 Pa s. Many salt layers are composed of rocksaltinterbedded with carbonate, clastic, or other evaporitic sedimen-tary rocks (e.g., Jordan and Nüesch, 1989; Jackson et al., 1990).Under similar conditions, the effective viscosity of an impure saltlayer probably exceeds that of a pure salt layer. Thus, availableinformation on fault displacement rates in nature and viscosities ofsalt layers indicates that P = drn • µn can range from about 105 to1012 Pa cm.

Using the above values for ρ*, g*, µm, l*, and P in equation 4,the displacement rate on the master normal fault in the modelscould range from 10-10 to 10-1 cm/s. In our models, the displace-ment rate varied from about 8.2 × 10-6 to 3.9 × 10-4 cm/s (Table 1).Models with low displacement rates simulated natural conditionswhere the value of P was low (i.e., the fault displacement rates innature were low or the evaporitic layer had a low viscosity).Models with high displacement rates simulated natural conditionswhere the value of P was high (i.e., the fault displacement rates innature were high or the evaporitic layer had a high viscosity). Eachmodel with its specified displacement rate represented a suite ofnatural conditions where P had the same value.

Restoration Methodology

We produced restorations for models 7, 8, 9, and 12 using thefollowing methodology: First, we created cross sections based onthe final structural configuration of the models. These final crosssections were compartmentalized into blocks bounded by faults orwith constant bedding dip. We selected a displacement value forthe master normal fault and identified the growth bed associatedwith that displacement value. The cross sectional blocks wererotated and translated until the growth bed became flat. We usedthe deformation patterns observed on the surface of the models,the known geometry of the fault blocks beneath the putty layer,and the assumption that the cross sectional areas of the overbur-den and putty remained constant during deformation to furtherconstrain the geometries of the restored cross sections.

We used a similar methodology to produce the restorations forthe Cormorant and Flying Foam structures of the Jeanne d’Arc Basin.We created regional cross sections based on seismic profiles dis-played with approximately no vertical exaggeration. These cross sec-tions were compartmentalized into blocks bounded by faults or withconstant bedding dip. We rotated and translated the cross sectionalblocks until the restored horizons became flat. We also assumed thatthe cross-sectional area of the suprasalt section remained constantduring deformation. We emphasize that these restorations areapproximations (i.e., we did not convert the seismic profiles todepth, and we did not decompact the sedimentary section).

REFERENCES CITEDBarrientos, J., 1989, Palinspastic fault reconstructions in the

central part of the Corsair trend, offshore Texas: Ph.D. thesis,Texas A&M University, College Station, Texas, 62 p.

Blystad, P., H. Brekke, R. B. Faerseth, B. T. Larsen, J. Skogseid, andB. Tørudbakken, 1995, Structural elements of the Norwegiancontinental shelf, part II: The Norwegian Sea Region,Norwegian Petroleum Directorate, Bulletin 8, 45 p.

Brace, W. F., and D. L. Kohlstedt, 1980, Limits on lithosphericstress imposed by laboratory experiments: Journal ofGeophysical Research, v. 85, p. 6248–6252.

Brown, R. N., 1980, History of exploration and discovery ofMorgan, Ramadan, and July oilfields, Gulf of Suez, Egypt, inA. D. Miall, ed., Facts and principles of world petroleumoccurrence: Canadian Society of Petroleum Geologists,Memoir 6, p. 733–764.

Bukovics, C., and P. A. Ziegler, 1985, Tectonic development ofthe mid-Norway continental margin: Marine and PetroleumGeology, v. 2, p. 2–22.

Byerlee, J., 1978, Friction of rocks: Pure Applied Geophysics, v. 116, p. 615–626.

Coffield, D. Q., and S. Schamel, 1987, Kinematics of drape foldevolution over normal faults in the Gulf of Suez, Egypt (abs.):Geological Society of America Abstracts with Programs, v. 19,p. 622.

Coffield, D. Q., and S. Schamel, 1989, Surface expression of anaccommodation zone within the Gulf of Suez, Egypt: Geology,v. 17, p. 76–79.

Colleta, B., P. Le Quellec, J. Letouzey, and I. Moretti, 1988,Longitudinal evolution of the Suez rift structure (Egypt):Tectonophysics, v. 153, p. 221–233.

Driscoll, N. W., J. R. Hogg, N. Christie-Blick, and G. D. Karner,1995, Extensional tectonics in the Jeanne d’Arc Basin, offshoreNewfoundland: implications for the timing of break-upbetween Grand Banks and Iberia, in R. A. Stoker, M. S.Shimmield, and W. W. Tudhope, eds., The tectonics,sedimentation and palaeoceanography of the North Atlanticregion: Geological Society Special Publication No. 90, p. 1–28.

Enachescu, M. E., 1987, Tectonic and structural framework of thenortheast Newfoundland continental margin, in C. Beaumontand A. J. Tankard, eds., Sedimentary basins and basin-formingmechanisms: Canadian Society of Petroleum Geologists,Memoir 12, p. 117–146.

Garfunkel, Z., and Y. Bartov, 1977, The tectonics of the Suez rift:Bulletin Geological Survey of Israel, v. 71, p. 1–44.

Handin, J., 1966, Strength and ductility, in S. P. Clark, Jr., ed.,Handbook of physical constants: Geological Society ofAmerica, Memoir 97, p. 223–289.

Higgins, R. I., and L. B. Harris, 1997, The effects of covercomposition on extensional faulting above re-activatedbasement faults: results from analogue modelling: Journal ofStructural Geology, v. 19, p. 89–98.

Holser, W. T., G. P. Clement, L. F. Jansa, and J. A. Wade, 1988,Evaporite deposits of the North Atlantic rift, in W. Manspeizer,ed., Triassic–Jurassic rifting, continental breakup and the originof the Atlantic Ocean and passive margins: Developments inGeotectonics 22, Elsevier, New York, p. 525–556.

Horsfield, W. T., 1977, An experimental approach to basement-controlled faulting: Geologie en Mijnbouw, v. 56, p. 363–370.

Hubbert, M. K., 1937, Theory of scale models as applied to thestudy of geologic structures: Geological Society of AmericaBulletin, v. 48, p. 1459–1520.

Jackson, M. P. A., R. R. Cornelius, C. H. Craig, A. Gansser, J. Stöcklin, and C. T. Talbot, 1990, Salt diapirs of the Great Kavir,central Iran: Geological Society of America, Memoir 177, 139 p.

Jacobsen, V. W., and P. van Veen, 1984, The Triassic offshoreNorway north of 62N, in A. M. Spencer et al., eds., PetroleumGeology of the North European Margin: Norwegian PetroleumSociety, London, Graham and Trotman, p. 317–327.

Jordan, P., and R. Nüesch, 1989, Deformational behavior of shaleinterlayers in evaporite detachment horizons, Jura overthrust,Switzerland: Journal of Structural Geology, v. 11, p. 859–871.

Koyi, H., and K. Petersen, 1993a, Influence of basement faults onthe development of salt structures in the Danish basin: Marineand Petroleum Geology, v. 10., p. 82–94.


Koyi, H., and K. Petersen, 1993b, Basement faulting—acomplementary triggering mechanism for diapirism: SEPMGulf Coast Section, 13th Annual Research Conference,Proceedings, p. 103–114.

Koyi, H., M. K. Jenyon, and K. Petersen, 1993, The effect ofbasement faulting on diapirism: Journal of Petroleum Geology,v. 16, p. 285–312.

Kukal, Z., 1989, The rate of geological processes: Earth ScienceReview, v. 28, p. 1–284.

Laubscher, H. P., 1982, Die Südostecke des Rheingrabens—einkinematisches und dynamisches Problem: Eclogae GeologicaeHelvetiae, v. 75, p. 101–116.

Maurin, J.-C., 1995, Drapage et décollement des séries jurassiquessur la faille de détachement majeure du rift rhénan sud:implications sur la géométrie des dépôts syn-rifts oligocenes:Comptes Rendus Académie des Sciences Paris, v. 321, p. 1025–1032.

Moustafa, A. M., 1975, Review of diapiric salt structures, Gulf ofSuez: Proceedings of the Ninth Arab Petroleum Congress,Paper 103 (B3), Abu Dhabi, 15 p.

Nalpas, T., and J.-P. Brun, 1993, Salt flow and diapirism related toextension at crustal scale: Tectonophysics, v. 228, p. 349–362.

Nicol, A., J. J. Walsh, J. Watterson, and J. R. Underhill, 1997,Displacement rates of normal faults: Nature, v. 390, p. 157–159.

Olsen, P. E., R. W. Schlische, and P. J. W. Gore, eds., 1989,Tectonic, depositional, and paleoecological history of earlyMesozoic rift basins, eastern North America: InternationalGeological Congress Field Trip T351, Washington, D. C.,American Geophysical Union, 174 p.

Olsen, P. E., R. W. Schlische, and M. S. Fedosh, 1996, 580 kyrduration of the Early Jurassic flood basalt event in easternNorth America estimated using Milankovitch cyclostratigraphy,in M. Morales, ed., The continental Jurassic: Museum ofNorthern Arizona Bulletin 60, p. 11–22.

Patton, T. L., 1984, Normal-fault and fold development insedimentary rocks above a pre-existing basement normal fault:Ph.D. thesis, Texas A & M University, College Station, Texas,164 p.

Patton, T. L., R. A. Nelson, A. R. Moustafa, and S. A. Abdine, 1994,Tectonic evolution and structural setting of the Suez rift, in S. M.Landon, ed., Interior rift basins: AAPG Memoir 59, p. 9–55.

Richard, P., and R. W. Krantz, 1991, Experiments on faultreactivation in strike-slip mode: Tectonophysics, v. 188, p. 117–131.

Robson, D. A., 1971, The structure of the Gulf of Suez (Clysmic)rift, with special reference to the eastern side: Journal of theGeological Society, v. 127, p. 247–276.

Rutter, E. H., 1986, On the nomenclature of mode of failuretransition in rocks: Tectonophysics, v. 122, p. 381–387.

Schütz, K. I., 1994, Structure and stratigraphy of the Gulf of Suez,Egypt, in S. M. Landon, ed., Interior rift basins: AAPG Memoir59, p. 57–96.

Sims, D., 1993, The rheology of clay: a modeling material forgeologic structures: EOS-American Geophysical UnionTransactions, v. 74, no. 46, p. 569.

Sinclair, I. K., 1993, Tectonism: the dominant factor in mid-Cretaceous sequence stratigraphy in the Jeanne d’Arc Basin,

Grand Banks: Marine and Petroleum Geology, v. 10, p. 530–549.Sinclair, I. K., 1995a, Transpressional inversion due to episodic

rotation of extensional stresses in Jeanne d’Arc Basin, offshoreNewfoundland, in J. G. Buchanan and P. G. Buchanan, eds.,Basin inversion: Geological Society Special Publication No. 88,p. 249–271.

Sinclair, I. K., 1995b, Sequence stratigraphic response toAptian–Albian rifting in conjugate margin basins: a comparisonof the Jeanne d’Arc Basin, offshore Newfoundland, and thePorcupine Basin, offshore Ireland, in R. A. Stoker, M. S.Shimmield, and A. W. Tudhope, eds., The tectonics,sedimentation and palaeoceanography of the North Atlanticregion: Geological Society Special Publication No. 90, p. 29–49.

Tankard, A. J., and H. J. Welsink, 1987, Extensional tectonics andstratigraphy of Hibernia oil field, Grand Banks, Newfoundland:AAPG Bulletin, v. 71, p. 1210–1232.

Tankard, A. J., H. J. Welsink, and W. A. M. Jenkins, 1989,Structural styles and stratigraphy of the Jeanne d’Arc Basin,Grand Banks of Newfoundland, in A. J. Tankard and H. R.Balkwill, eds., Extensional tectonics and stratigraphy of theNorth Atlantic margins: AAPG Memoir 46, p. 265–282.

Thiebaud, C. E., and D. A. Robson, 1979, The geology of the areabetween Wadi Wardan and Wadi Gharandal, east Clysmic rift,Sinai, Egypt: Journal of Petroleum Geology, v. 1, p. 63–75.

Thiebaud, C. E., and D. A. Robson, 1981, The geology of the Asloilfield, western Sinai, Egypt: Journal of Petroleum Geology, v. 4, p. 77–87.

Tsuneishi, Y., 1978, Geological and experimental studies onmechanism of block faulting: Bulletin of the EarthquakeResearch Institute, v. 53, p. 173–242.

Urai, J. L., C. J. Spiers, H. J. Zwart, and G. S, Lister, 1986,Weakening of rock salt by water during long-term creep:Nature, v. 324, p. 554–557.

Vendeville, B., 1987, Champs de failles et tectonique enextension: modélisation expérimentale: Ph.D. Thesis,Université de Rennes, Rennes, France, 395 p.

Vendeville, B., 1988, Modèles expérimentaux de fracturation delat couverture controlée par des failles normales dans le socle:Comptes Rendus Académie des Sciences Paris, v. 307, série II,p. 1013–1019.

Vendeville, B., G. Hongxing, and M. P. A. Jackson, 1995, Scalemodels of salt tectonics during basement-involved extension:Petroleum Geoscience, v. 1., p. 179–183.

Weijermars, R., 1986, Flow behavior and physical chemistry ofbouncing putties and related polymers in view of tectoniclaboratory applications: Tectonophysics, v. 124, p. 325–358.

Weijermars, R., M. P. A. Jackson, and B. Vendeville, 1993,Rheological and tectonic modeling of salt provinces:Tectonophysics, v. 217, p. 143–174.

Withjack, M., K. Meisling, and L. Russell, 1989, Forced folding andbasement-detached normal faulting in the Haltenbanken area,offshore Norway, in A. J. Tankard and H. R. Balkwill, eds.,Extensional tectonics and stratigraphy of the North Atlanticmargins: AAPG Memoir 46, p. 567–575.

Withjack, M., J. Olson, and E. Peterson, 1990, Experimentalmodels of extensional forced folds: AAPG Bulletin, v. 74, p. 1038–1054.


Martha Withjack

Martha Oliver Withjack received her Ph.D. from BrownUniversity in 1978, studying the mechanics of continentalrifting. Before joining Mobil Research and DevelopmentCorporation in 1988, she worked as a research geoscien-tist at Cities Service Oil and Gas Company and ARCO Oiland Gas Company. Her research interests include exten-sional, inversion, and salt tectonics, physical and analyticalmodeling of structures, and structural interpretation ofseismic data. She was an AAPG Distinguished Lecturer(1984–1985), a recipient of the J. C. “Cam” SprouleMemorial Award (1986), a Distinguished Lecturer for thePetroleum Exploration Society of Australia (1996), a recip-ient of the Matson Memorial Award (1999), and a fellow ofthe Geological Society of America. She is now a professorof geology in the Department of Geological Sciences,Rutgers University.

Sybil Callaway

Sybil Callaway received her B.A. degree in geology fromLouisiana State University (1979) and her M.S. degree ingeology from the University of New Orleans (1989). Sheworked from 1981 to 1985 with Texaco, Inc., as an engi-neering geologist in Morgan City, Louisiana, and from1986 to 1990 as an engineering geologist with two inde-pendent oil companies in New Orleans. In 1991, shebegan work as an exploration geologist for MobilExploration Producing, U.S., Inc., where her work on sub-salt prospects lead her to develop structural expertise inaffiliation with Mobil Exploration and ProducingTechnology Center (MEPTEC) in Dallas, Texas. Sheworked with MEPTEC from 1994 to 1998. Currently, sheis working as a production geologist for Mobil ProducingNigeria, in Lagos, Nigeria.


ABOUT THE AUTHORS

Documents

Active Normal Faulting Beneath a Salt Layer: An ...schlisch/aapg_ff_2000.pdfdevelopment of extensional forced folds above ... extensional forced folding and within the underly-