GeoCh4

C O N T E N T S

1. PURPOSE AND SCOPE

2. STRUCTURAL FEATURES

3. RELATIONSHIP OF STRUCTURES TOGEOLOGICAL EVENTS

4. PRACTICAL STRUCTURAL GEOLOGY

5. FURTHER LEARNING

4Structural Geology4

2

LEARNING OBJECTIVES:

At the end of this Chapter the student will be able to:

• Give names for the main types of structural features

• Identify the characteristics of a structural trap

• Describe the main types of faults, and indicate how these structures relate to thetectonic setting

• Describe typical fold geometries found in the main tectonic settings, and indicatehow these may relate to other structural features

• Understand the importance of fractures and other localised deformation in terms oftheir impact on fluid flow characteristics

• Describe the fracture patterns associated with folding

• Identify structural features in core and on dipmeter/image logs

• Identify structural features on cross sections, maps, and interpreted seismic lines

• Describe the causes of, and identify, situations leading to fault compartmentalisation

Department of Petroleum Engineering, Heriot-Watt University 3


1.PURPOSE AND SCOPE

Structural Geology is a component of Petroleum Geology. It is closely linked toGeophysics and Rock Mechanics, and there also are strong links to Sedimentation.Structural Geology has historically been concerned with the study of rock deformationover geological timescales - sometimes, this activity is also referred to as Tectonics.Modern Structural Geology has evolved to include an appreciation of Rock Mechanics,making it possible to produce predictions about the conditions of deformation, andabout the impact of that deformation on rock properties. Rock Mechanics alsoincludes a consideration of human timescales, and, particularly, how human activitycan influence rock deformation, and vice versa.

In the context of Petroleum Geoscience, Structural Geology is concerned withcreating an interpretation of the deformation history of a zone of interest in thesubsurface in order to understand the trapping potential (exploration) and/or howproduction may be affected by the deformation. Traditional structural geologists haveoften focused on an analysis of surface outcrops, where the observable scales ofdeformation and techniques are usually quite different from those applicable to thesubsurface. Structural geologists working in a petroleum setting have had to learn tosynthesise a wide variety of ideas drawn from many scales and many types ofstructures. This review of Structural Geology focuses on those concepts andtechniques that are most-applicable at the reservoir scale, making Structural Geologyrelevant to the needs of petroleum engineers.

To meet this goal, this Chapter will address:

(a) The role of Structural Geology in exploration:

• Structural trapping mechanisms - faulting, folding• Structural history - burial, tectonism and uplift• Structural features on seismic lines, wireline logs and cores

(b) The role of Structural Geology in reservoir development:

• Fracture production• Fault properties - compartmentalisation, sub-seismic faulting• Fault maps - juxtaposition diagrams, shale gouge ratio• Stress effects

The Chapter begins by describing the geometries of important structures and theassociated terminology. The next section discusses the relationship of structuralfeatures to other geological events. The following section focuses on practical issuesinvolving applications of the concepts presented previously. A final section suggestssome ways to learn more about this subject, including a set of exercises. RockMechanics concepts are treated in an Appendix.

Because of the close links between Structural Geology and Rock Mechanics, thisChapter and the Appendix should be studied in parallel.

4

2. STRUCTURAL FEATURES

2.1 FaultsA fault is a more-or-less planar surface or zone, across which the rocks on either sidehave been moved by shear displacement (i.e. displacement parallel to the faultsurface). Faults can be sharp (infinitesimally-thick) planes, and they can also be widezones consisting of an array of complex deformation features. Faults represent ayielding of the rock mass, and importantly, they indicate that deformation has becomelocalised (as opposed to distributed) at the scale of observation. Fault geometries, andfault patterns, are used to infer large-scale deformation states.

The majority of faults are not vertical; instead, most are inclined. The angle “down”from horizontal is called the dip of the fault plane, and the compass direction of thehorizontal line lying in the fault plane is called the strike (Fig. 1). A vertical fault hasa dip of 90o, and non-vertical faults have dips that range from very shallow (10-30o)to moderate (40-60o) to steep (70-89o). The dip of the fault plane, along with the senseof motion (see below), is used to categorise the types of faults.

Left-Hand Rule:If left thumb points down dip,then left index finger points in strike direction.

In this example: dip = 22º, strike 105º

This is written as: 22/105(other conventions exist)

��

yyyyyy��yyy

��

yyyyyyyN

75ºStrike

Dip

300m

Horizontal Line

Another Rock Layer

Angle of Dip (=22º) Top of Rock Layer

200m

100m

0m

Note: Structural Geology is intimately linked with geometries. The definitions andexplanations of structural features necessarily require drawings of these geometries.There are three principal methods used to illustrate structural forms (Fig. 2). The 3-D block diagram is perhaps the most readily understood of the three methods, sinceit is “visual”. A cross section can be thought of as being the side of a block diagram(even if the other side and top are not shown), and a map is simply the top of thatdiagram projected onto a 2-D plane (a piece of paper). Although there are variouskinds of maps (see Chapter 7), structural features are usually described using eithergeological maps or structure contour maps.

Figure 1

Definition of dip and strike

of a surface



Geological Map

Cross Section

Block Diagram(Showing a Dome)

��

yyyyyyyyyyyy��

yyyyyyyyy�y

��yyyy

��yy�y��yy

��

��

��

yyy

yyyy

yyyy

��yyy��yyyy

��yyyy��yyy

��yy

��

yyyyyyyyyyyy

��yy

��yyyy

��yyyyyy

��

yyyyyyyyyyyyyyy

��

yyyyyyyyyyyyy��yyy

��

yyyy

��

yyyyyy

��

��

��

��

yyy

yyyy

yyyy

yyyy

��

��

��

��

��

yyy

yyyy

yyyy

yyyy

yy

��yy

��

��yy

yy

��

��

��

yyy

yyy

yy

Top of Block

Side of Block

Each fault cuts the entire rock mass into two fault-blocks. In the case of non-verticalfaults, the fault-block lying below the fault plane is called the footwall, regardless ofthe sense of displacement of the fault, and the block above the fault is called thehangingwall (Fig. 3). The terms footwall and hangingwall derive from the miningindustry where fault planes are often encountered in underground workings (somefaults in orogenic belts are subject to mineralisation, implying that they were goodfluid conduits). The footwall (imagine a miner’s feet on the footwall) and hangingwall(imagine a miner hanging from the roof of the mine) cannot be defined for a verticalfault.

Footwall

Footwall

Hangingwall

Hangingwall

Normal Fault Reverse Fault

The types of faults are defined by the sense of movement along the fault plane (Fig.4). In dip-slip faults (the slip motion is parallel to the dip direction), if the hangingwallmoves down (with respect to the footwall), this is called a normal fault. Normal faultsare associated with extension (lateral increase in dimension). If the hangingwall risesover the footwall, this is called a reverse fault. Reverse faults (thrusts are reversefaults whose dip is low - less than 25o) are associated with shortening (lateral decreasein dimension). In the case of strike-slip faults (where the movement is parallel tostrike), we use the terms left-lateral and right-lateral to indicate the sense of relativemotion as seen looking down on a map.

Figure 2

Types of illustrations used

to depict structural features

Figure 3

Definition of the terms

footwall and hangingwall,

with respect to the dipping

fault plane, but independent

of the sense of movement

6

NORMAL

REVERSE

STRIKE SLIP

Lateral Movement

extension

contraction

Faults can be simple or complex zones of displacement (Fig. 5). In the petroleumindustry, faults are often represented as simple, single breaks (this is almost always thecase at the scale of a reservoir map), but they are in reality more likely to be complexzones. A single fault break may well become a much more complex fault zone as thescale of observation magnifies. Large, “intact” blocks of rock that are found withina fault zone, surrounded by sheared and distorted rocks, are called “horses”. Synthetic(similar dip) and antithetic (opposite dip) faults are names applied to minor faults thatare associated with larger faults.

Faults tend to form in groups, or arrays, giving rise to various geometric arrangementsthat have the potential for trapping hydrocarbons. For an array of faults that have asimilar strike, it is common to find that some of the faults dip in one direction, whilesome dip in the opposite direction (see Rock Mechanics for an explanation). Whenseen in a cross-sectional view (Fig. 6), such an arrangement of normal faults producesblocks that are uplifted or dropped down relative to one another. In this pattern, theuplifted blocks are called horsts, and the down-dropped blocks are called grabens.Curiously, deformations that shorten the layers seem to be less likely to produce equalnumbers of left-dipping and right-dipping faults; perhaps this is related to the idea thatsuch faulting is related to the spreading of a crustal-scale wedge of rocks, a processwhich has a preferred direction.

Figure 4

Types of fault defined by

displacement along the

fault plane A: Normal (dip-

slip) fault; B: Reverse (dip-

slip) fault; C: Strike-slip

fault (left-lateral shown)

Figure 5

Types of fault zones. A: a

simple, single surface of

shear; B: a fault zone

composed of a set of shear

surfaces; C: a distributed

(ductile) shear zone



Horst

Thrust Fault (Listric)

Horst

Graben

Half Grabens

Listric FaultAntithetic

Synthetic

EXTENSION

CONTRACTION

Large faults are often not planar. A listric fault is a curved fault that is steeply-dippingat shallow structural levels, and gently-dipping at deeper levels. Listric normal faultsare important in extensional domains because they provide for rotations of fault blocks(Fig. 7). The tilting can provide a chance for new sediment deposition, which, in thiscase, produces a wedge-shaped rock succession. Such a situation is called a half-graben (it is a down-dropped block, but only on one side). The shallow portion (crest)of a tilted fault block is a potential hydrocarbon trap, while the deep portions of large,tilted fault blocks may subside far enough to become heated and thus turn into ahydrocarbon kitchen. Listric faults are also important in shortening, where they maybe associated with folds (see below).

early-matured hydrocarbons

shales

unconformity

source rock

source rock

��yy��yy��yyyyyy

compaction

Reservoir Unit

Small hydrocarbonaccumulations

migration along fault?

coarse clasticsfrom erodedfault-block crest

source rock

source rock

Reservoir Unit

late-matured hydrocarbons

erosion

migration

sandstones leaks, re-migrated hydrocarbons, re-trapped in shallow layers

migration

early-trappedhydrocarbons

late-trappedhydrocarbons

Fault displacements are not constant along a fault surface. Displacement variessystematically along a fault and reaches a maximum near the centre of the fault

Figure 6

Sketch of A: extensional;

and B: shortening

structural regimes

Figure 7

Importance of extensional

faulting relative to the

Petroleum System

8

surface. The edges (or limits) of a fault surface, known as the fault tips, are locationsof zero displacement. Real fault zones are usually made up of numerous faultsegments that partly overlap, with intervening regions of distorted rock (refer to Fig. 8).

��yyy��yy��yy�y��yy��yyy

��yyyyyyyy

��yy�y�y��yyyy

��yyyyyy

��yyyyyy

Complex Fault Zone Detail of Distorted Rocks and Overlapping Fault Segments

It has been observed that the larger (longer on a map or cross section) faults in a systemhave the largest displacements. It is also noted that there tend to be more, smallerfaults, and fewer, large faults. This relationship is often expressed as a power-lawcurve (Fig. 9). With such a relationship, the sub-seismic faults (those that are toosmall to be identified on seismic data) can be modelled (but see below for cautionsagainst using this technique blindly).

1 10 100 1000

0.01

100

0.1

10

1

Maximum Throw (m)

Cum

ulat

ive

faul

t den

sity

(/k

m2 )

"Sub seismic"

"Seismically Resolvable"

The common fault types are sometimes interpreted to imply the orientation of theprincipal stresses that existed when the fault was formed (see Appendix). In thisapproach, ideal normal faults would have the maximum compressive pricipal stress(σ

1) vertical, ideal reverse faults would have the minimum principal stress (σ

3)

vertical, and ideal strike-slip faults would have the intermediate principal stress (σ2)

vertical (Fig. 10). Although this simple relationship is a useful learning tool, real statesof stress are rarely as uniform and homogeneous as implied by this conceptual model(this issue is treated further in the Rock Mechanics Appendix), and its over-application is a common pitfall.

Figure 8

Examples of fault zone

complexity

Figure 9

Fault population curve

exhibiting a power-law

relationship. Such curves

have been used to

extrapolate fault

populations below the limit

of seismic resolution



σ1

σ2

σ3σ1

σ2

σ3

σ1

σ2

σ3

30º

2.2 FoldsA fold can be defined as the deflection of a marker surface (e.g. from a planar shapebefore folding to one that is non-planar after folding). The majority of situations ofinterest to Petroleum Engineers involve layered, sedimentary rocks, so the “markersurface” is usually bedding (but this is not always the case). Even if the rocksuccession is a monotonous stack of the same layer on top of the same layer, thebedding planes represent a significant element of heterogeneity. In the more usualcase where there are lithological variations (e.g. sand/shale/sand, etc), there is an evengreater degree of heterogeneity. These mechanical variations are what make “fold-ing” a distinctive structural process.

You will probably be surprised if you consult a range of textbooks to get another pointof view concerning folding: none give a definition of folding that is any more definitethan that given in the preceding paragraph! The reason for this “hedging” is that mostgeologists wish to classify folds according to their mode of origin: i.e. their genesis.If it were possible to do this un-ambiguously, then the problem would be solved.Unfortunately, folding processes are still the subject of research, and the genesis ofany particular fold cannot be determined with certainty, so this strategy cannot beadopted. We will return to the topic of fold genesis, but first, we need to be sure thatwe know the terminology of fold shapes.

Fold shapesThe first point to make concerns the word “surface”. Rock layers, and other-shapedrock bodies, are bounded by surfaces. A surface is a curvi-planar entity that, inmathematical jargon, has only two dimensions (a surface has infinitesimal thickness).A large portion of the fold-shape naming scheme (Fig. 11) is based on the shape ofsurfaces. In this context, the “surface” is usually the bedding plane bounding the topof a rock layer. (But, remember, this bedding plane also bounds the bottom of the next-higher layer.) The key words are: crest line (the line that represents the locally-highestelevation), trough line (locally-lowest elevations), inflection line (boundary betweenconvex-upwards and convex-downwards), culmination (highest point of a crest ortrough line), and depression (lowest point of a crest or trough line). Note that the shapevariations of surfaces are distinctly three-dimensional (it is important that you don’tforget that the 2-D cross-section drawings of folds that we use are illustrative, and notvery realistic!).

Figure 10

Orientation of principal

stresses for the three ideal

types of faults

10

Curvi-Planar Surface

Culmination Depression

c

c

c

i

t

t

ii

i

ii

Crest Lines (c), Trough Lines (t) and Inflection Lines (i)

Much of the following material is based on how a cross section of a surface appears.In mathematics, the key concept needed for evaluating the shape of a surface iscurvature. Curvature is actually the inverse of the radius ( 1 / r ) of the circle whichhas the same shape as a small segment of the surface. If we determine the curvatureof a folded surface (via equations that are not important for our purposes), there willusually be some places where the curvature is higher than in nearby regions (i.e. thefolding of the surface is “tighter” in the high-curvature sites). We call such high-curvature parts of a fold “hinges”. The less-curved (straighter) portions of the foldare called “limbs”. Both single-hinge and multi-hinge folds are possible. The foldaxis is the line that is formed from the intersection of the axial surface with some layerboundary. The fold axis and the hinges are usually parallel (Fig. 12).

High-curvature sites(Hinges)

Less-curved regions(Limbs)

Stylised Representation withStraight Limbs and Point Hinges

InterlimbAngle

A useful geometric characteristic of a fold shape is its interlimb angle (Fig. 12). Thisparameter has to do with the apparent tightness of the fold, or the angularity of thehinges. The following table gives the descriptive terms that are applied to ranges ofthe measured interlimb angle.

Figure 11

Terms that describe the

shape of a curvi-planar

surface

Figure 12

Fold limbs and hinges



Descriptor Interlimb Angle (deg)

Gentle 180 - 120

Open 120 - 70

Close 70 - 30

Tight 30 - 0

Isoclinal 0

Other terms are used to describe fold patterns, including fold symmetries. Althoughall folds do not occur in the form of wavetrains, the following definitions are mosteasily visualised if we draw such a set of repeating fold forms. Enveloping surfacesdelimit the deflections of the surface about some median position. The fold amplitudeis half the distance between the enveloping surfaces. The median surface joins theprimary inflection points. The wavelength is the distance between comparableinflection points. Note that the wavelength is shorter than is the distance betweeninflection points measured along the layer. Symmetric folds have equal-length limbs,and asymmetric folds have unequal-length limbs (Fig. 13).

A

A

As As

AsW

i i i i Median Surface, m

Periodic Symmetrical Waves

Periodic Asymmetrical Waves

W = WavelengthA = Amplitudei = Inflection PointsAs = Axial Surfaceθ = Inclination of axial

surface relative to enveloping surface

iii i m

W

As

As

As

θ

Enveloping Surface

Enveloping Surface

Figure 13

Terms applied to repeating

fold shapes

12

Of course, our real interest is in the folding of rock layers (which have finitethicknesses). When a layer is flexed “up”, we call this structure an anticline; whenthe layer is flexed “down”, we call this structure a syncline (Fig. 14). Becausepetroleum is usually more buoyant than the aqueous fluids that are otherwise presentin the pore spaces of rocks, it tends to migrate upwards. An anticline is an ideal shapethat could serve as a trap for oil and gas (assuming that the migrating hydrocarbonshad entered into the reservoir layer, had moved along it, and that overlying it there wasa seal; see separate Chapter on Petroleum Play).

A

Anticline: bows upwards, older rocks inthe middle, dips are "Away"

Syncline: bows downwards, younger rocksin the middle, dips to the centre

Our real interest is in the (usual) case where there are stacks of layers. Imaginarysurfaces can be created (Fig. 15) to join up the various points defined on each layerboundary. Axial surfaces bisect the limbs. For the types of folds usually encounteredin petroleum systems (e.g. approximately-constant layer thicknesses; see below), theaxial sx"Vace is essentially the same as the hinge surface, which joins all of the hingepoints. Inflection surfaces join all of the inflection points, and crest and troughsurfaces (not usually drawn) could join all of the high and low points, respectively.Inflection surfaces define fold domains - within which the changes of fold shapes areusually regular and predictable.

Cross section of multi-layer stack showing complex array of axial surfaces

��

yyyyyyyyyyyy

Inflection Surfaces Define Fold domains

Figure 14

Definition of anticline and

syncline

Figure 15

Imaginary surfaces

segmenting a folded

succession



A fold may well have a fold axis that is not horizontal. Non-horizontal axes are saidto be plunging (Fig. 16). (A plunging line is specified by its trend - compass directionof the downward-pointing end of the line, and its plunge angle - measured down fromhorizontal). If a fold has a plunging axis, and a non-vertical axial surface, its axial trace(where the axial surface intersects the ground), and its crestal trace (where the crestalsurface intersects the ground), may not coincide on a map. This is a point that seemsto greatly trouble many students! The fold profile is a section of the fold taken at rightangles to the fold axis. In plunging folds, the profile is not the same as the (vertical)cross section.

True Shape (Profile) of Plunging Cylinder

Cross Section of Pipe

Map Shape

Plunge Angle

Fold Axis

Horizontal Line

Axial Surface

One fold classification scheme is based on variations or constancy in the thickness ofthe folded layers (Fig. 17). Parallel folds have constant layer thicknesses (measuredperpendicular to the layer boundaries). Concentric folds are special parallel folds thathave more-or-less constant curvature centered on a point. Similar folds have aconstant distance between layers (measured along a direction that is parallel with theaxial surface).

Figure 16

A plunging fold

14

t

t t

t

Parallel FoldThickness (t) constant along layer

Concentric (Parallel) FoldConstant layer thicknessand constant curvature

Another classification scheme relies on the patterns of dip isogons as viewed in a foldprofile (Fig. 18). Isogons are lines joining points on different surfaces that have thesame dip. There are three classes, although the first class is sub-divided into 1A, 1B,and 1C sub-classes. In a multi-layer sequence of rocks, and especially if there arestrong mechanical contrasts between the layers, mixed fold classes can occur. Mostof the folds of interest to Petroleum Engineers belong to Class 1B.

1A

1A1B Parallel

1B Parallel

Class 2

Class 3, Divergent Isogons

1C

1C

2 Similar

3

Fold Shape (Classes)

Dip Isogons (lines of constent dip)

A suite of terms exists to provide succinct communication about the orientation of afold (Fig. 19). These terms are based on: the fold’s tightness (i.e. its interlimb angle;see above); its symmetry; the dip of the axial surface; and the plunge of the fold hinge.

Figure 18

Fold classification based on

dip isogons

Figure 17

Parallel and concentric

folds



Descriptor Dip of Axial Surface (deg)

Horizontal, Sub-horizontal 0, 1 - 10

Gently Inclined 10 - 30

Moderately Inclined 30 - 60

Steeply Inclined 60 - 80

Sub-vertical, Vertical 80 - 89, 90

Descriptor Plunge Angle (deg)

Horizontal, Sub-horizontal 0, 1 - 10

Gently Inclined 10 - 30

Moderately Inclined 30 - 60

Steeply Inclined 60 - 80

Sub-vertical, Vertical 80 - 89, 90

Each combination of terms suggests a geometric image. A full description links theseterms together: e.g. an asymmetric, tight, inclined, plunging anticline. If thedescriptors are quantified (e.g. strike and dip of the axial surface, trend and plungeangle of the hinge, etc), then the fold orientation is also communicated.

Horizontal Fold Axis

Upright Horizontal

Inclined Horizontal Inclined Plunging

Recumbent Horizontal

Upright Plunging

Vertical Fold Axis

Vertical

Plunging Fold Axis

Figure 19

Fold orientations

16

In practice, these terms are not frequently used by petroleum geoscientists orengineers, but it is useful to know of their existence so that you will be able tounderstand fold descriptions that you might read, or that might arise in specialcircumstances. Although the terms are not frequently used, it is worth noting that, ifa fold can be described in this way, there is an implication that the fold’s geometry issomewhat “regular”, and hence, predictable.

Types of flexureMany geologists make a distinction between buckle folds and folds created bybending. Buckles are supposed to result from the shortening of a layer, while bendsare created by displacements imposed normal to the layering. However, bending may(and usually does) produce folds that are shorter (occupy less distance on a map) afterfolding than the rocks were before the folding, so this view is inadequate to distinguishbetween these types of fold.

A slightly more accurate notion about buckling is the one that states that the rock layeris “pushed” along its length (Fig. 20). In practice (for natural folds), this statementmeans that distant points on the layer are pushed closer together, and the layer deflects(folds) away from its previous planar shape. The shape of the deflections (primarilytheir wavelength) is controlled by a number of factors, including the thickness of thelayer, and its properties relative to the properties of the surrounding materials. Thissort of fold model is thought to produce a wavetrain of folds that have similargeometric characteristics. Individual folds in the wavetrain grow by increasing theiramplitude, and, because the fold hinges are assumed to be “fixed” in the rock, thewavelength decreases simultaneously.

Loading Examples

Buckling

Bending

Fold Train

Block Fault

DifferentialCompaction

Diapir

In map view, buckle folds are associated with a strain ellipse (refer to Rock MechanicsAppendix) that has its short semi-axis oriented in the “push” direction, and its longsemi-axis oriented along the fold axes. Most real buckle folds are not infinite alongtheir axis, but instead transfer their shortening to neighbouring folds. The length-to-width ratio of most buckle folds (as seen on a map) is about 10:1.

It is appropriate here to introduce a notion that applies to all sorts of structures, but

Figure 20

Differences in loading: (a)

buckle folds and (b)

bending



which is particularly well illustrated by the image of fold wavetrains. As noted above,folds tend to be elongated. In a map view, this orientation of this elongation is referredto as the longitudinal direction, or the trend. These words are also used to indicate the“long” direction of any structural type. The perpendicular orientation on the map iscalled the transverse direction, or, less-precisely (but more commonly), the “dipdirection”. Although cross sections can be constructed along any alignment, it is mostcommon to draw them in the transverse or “dip” direction, since this orientationillustrates most-effectively the major changes in shape. For specific purposes, othercross-section alignments can be used.

A point not (yet) recognized in the literature is that, in a stack of layers, with eachhaving its own mechanical properties, one layer might buckle whilst the others aresubjected to bending because of the deflection associated with the controlling buckle.In regions that have been shortened, there are important genetic associations betweenthrust faults and bedding-plane faults, and folds (see later in this Chapter).

Regardless of the cause of the folding, flexures in the upper crust of the earth (i.e. thoseof interest to Petroleum Geoscience) are highly dependent on the fact that the rocksare layered. The bedding planes between rock layers are mechanical discontinuitiesthat are available for slip. When folding occurs, some (but not all) bedding planes doslip, and this process (called flexural slip folding) dramatically alters the pattern ofdeformation - as compared to the folding of a single thick layer that does not haveinternal bedding planes. Bedding-plane slip is extremely important in terms oflimiting the magnitude of the strains that are created at any point in the flexedsuccession of layers (e.g. fracture intensity), and in controlling the extent of fracturesthat may be induced. When flexural slip occurs (this is the “normal" case for the upper-crustal flexures of interest to Petroleum Engineers), the fractures that are created aredistributed differently than would be predicted by power-law relationships (seeabove). Because of the partitioning of strain associated with flexural slip, sequencesof rock layers can be considerably bent without undergoing extreme internal distor-tion (Fig. 21).

A.

B.

C.

D.

ActiveSlipSurfaces

Bending Strains (expressed as fractures)

To-be-ActivatedSlip Surface

Now-ActiveSlipSurfaces

New Bending Strains

ResultingSuperposedFracture Strains

Figure 21

Deformation in flexural-

slip folds

18

2.3 Fault / Fold Interactions

Contractional structuresThere is a major chicken-and-egg question that arises in the interpretation of folds andthrusts. It is common in fold/thrust belts (see below) to observe large-scale,asymmetric folds whose overturned limbs are faulted (Fig. 22). Is the faulting aconsequence of the folding, or is the folding a consequence of the faulting?

?

?

100's of m

Reverse faultcutting overturnedforelimb

It has been possible to develop kinematic models of the development of folds that arerelated to fault movements. The various types of model include: fault-bend folds,fault-propagation folds, detachment folds, and break-thrust folds. Now, some twentyyears into this process of creating these kinematic models, they have become verysophisticated (but not necessarily right!). There are models that address variations inlimb dips, interlimb angles, fault dips, displacement gradients on faults, fixed versusmoving hinges, layer thickness changes, etc, etc. The hope has been that differentmodes of formation could be distinguished on the basis of differences that could -ostensibly - be measured in a given structure. Unfortunately, clever people keepfinding a new parameter that allows them to develop another model to permit them tointerpret their favourite fold as being of type X. Although the initial motivation - todistinguish the mode of formation - may not be met, these kinematic models are usefulin that they provide us with hypotheses about how these types of structures form andevolve.

Some important aspects of these kinematic models of contractional structures are (Fig. 23):

• The (oft-assumed) ramp-flat-ramp shape of thrust faults• The footwall cut-off angle• The angle between the bedding that is being truncated and the hangingwall flat• Whether folds remain as the parallel type of whether the layer thicknesses change• Whether fault displacement is constant• Strong potential for syn-deformation deposition of sediments, and their immediate

involvement in further deformation

Figure 22



Ramp

h

tf

fα

2θ

Flat

FlatShape of Fault

Fold geometry produced by motion along fault

γ

In recent works, various researchers use the kinematic models to calculate an apparentstrain state in the resulting structures. If the inferred strains are thought to relate todeformation (e.g. fractures), then there is potential to use such models to aid inreservoir management. However, we do not yet have good forward simulations thatare based on proper rock mechanics, so any such predictions about the mechanicalstate should be treated cautiously. Probably the most crucial shortcoming of theseefforts is that none (to date) consider the bending of the layers, or the impact offlexural-slip folding.

Extensional structuresIn contrast to the situation with fold/thrust structures, there has been considerably lessattention paid to fault/fold relationships in extensional settings. Although thisstatement is true, a major example of such a relationship has been known for a verylong time: roll-over anticlines (Fig. 24). However, the kinematics of such featureshave been studied and understood for only a short time.

Note truncations

Rolloveranticline

listric normal fault

(tracing fromseismic image)

Figure 23

Figure 24

20

In extensional deformation, the orientation of most faults is at high angles to thelayering. This means that any resulting folds are ‘bends’, and certainly not ‘buckles’.Until fairly recently, it was widely believed that folding in extensional situations didnot occur, and that, instead, the rock layers simply faulted. (It was believed that thelayers would be stretched, and that the resulting state of stress would have very lowstress magnitudes, leading to a situation where the rocks were very weak, and henceliable to fault. This belief failed to account for the operation of the whole system, andespecially for flowage of some of the rock units.) Any observed flexure of the layeredrocks was dismissed as ‘drag’. Fortunately, good research is now showing thatextension-related folding is an important structural style that can be rationallyexplained using similar kinematic approaches as have been used for contractionalsettings. Full forward simulations are, however, not yet available for these structures,either.

Some important aspects of these kinematic models of extensional structures are:

• Angle of layering relative to local fault orientation (many extensional faults are listric)• Constancy / variability of displacement on the fault• Presence/absence of ‘weak’ units to permit flow, and related detachment of

overlying layers• Strong potential for syn-deformation deposition of sediments, and their

immediate involvement in further deformation

2.4 Fractures and JointsWhen rocks become deformed, they exhibit one or more of a variety of yield responses(see Appendix). One of the most common types of response is fracturing, or breakageof the rocks. Each such break represents a mechanical discontinuity in the rock mass.These structural discontinuities are amongst the most common of all geologicalfeatures: every outcrop and most cores exhibit some sort of fracturing. Fractures andother discontinuities affect nearly every petroleum reservoir, either by enhancing theproduction, or by causing problems for production.

There are different terms that can be used to refer to structural discontinuities —roughly distinguished by the scale of the feature and the amount of displacement.Although other authors may propose very specific definitions, we recommend apragmatic approach (Fig. 25). We suggest that the terms fracture (preferred) or jointbe used for sharp, localised breaks or discontinuities, and that the term fault be usedto refer to a plane or zone across which there is considerable shear (relativedisplacement parallel to the plane/zone). There is inevitably a difficulty in judgingbetween the use of the terms fracture or fault in some cases. (Many geoscientists wishto use the term joint for fractures with little discernable movement, but any supposeddistinction between the terms joint and fracture is of little or no practical use.)



��yyy��yyy��yyy��yyy

Fracture(Joint)

Fault

Same, Unspecified Scale

Fractures and/or joints often occur in systematically-aligned groups such that there isa similar dip and strike to each of the fractures in the group. Such a group is calleda fracture set or joint set (Fig. 26), and sometimes the word “ systematic” is added (e.g.systematic fracture set). Multiple sets of fractures/joints (each set being characterisedby a different strike and dip) also commonly occur; these groupings of sets are calledfracture assemblages if they are thought to be causally (genetically) related.

��yyyyyy

��yyyyyy

��yy��yy��yy

��yyyyyy

��

yyyyyyyyyyyy

��yy��yy��yy

A Fracture Set Two Related Fracture Sets(= Fracture Assemblage)

Fractures

There is presently a considerable level of interest in fractures. The primary stimulusis to understand the impact of fracture patterns on reservoir performance. A typicalfracture analysis will attempt to determine:

• the distribution and geometry of the fracture system(s)• the surface features of the fractures• the relative timing of the formation of different fractures• the geometric relationship of fractures to other structures

This knowledge can lead to the creation of a genetic model for the formation of thefractures, from which predictions can be made concerning the distribution in areas thatcannot be sampled.

The typical spacing between fractures in a set is often seen to be a function of lithologyand bed thickness (Fig. 27).

Figure 25

Displacement distinguishes

between fractures and

faults

Figure 26

Illustration of fracture sets

and assemblages

22

Spacing Between Fractures (m)

Limestones

�Sandstones

Bed

T h

ickn

ess

(m)

1�� 2

0.6

1.2

Features on the surface of a fracture can provide important information about its origin(Fig. 28). A set of curvilinear ribs, defining a feather-like (“plumose”) structure, isevidence of an extensional fracture (Mode I; see Appendix). Mineralised surfaceswith lineations are called “slickensides”, and these are indicative of shear movement(usually Mode II). Crystal growth on the surface of a fracture is an indication that thefracture has been a void (open space) at some point in its history, allowing mineralsto grow from circulating fluids. A special case of fracture-surface mineralisation isthe “crack-seal” arrangement, from which we can infer that mineral deposition wasconcurrent with fracture opening. In carbonate rocks, fractures can become zones ofdissolution, leading to open fissures. The nature of fracture surfaces is an importantconsideration for the performance of the fractures during production of a reservoir.Partly-mineralised, uneven fracture surfaces are less likely to close up (as fluid iswithdrawn) than may be the case with simple, co-planar extensional fractures.

Fracture Surface

Direction of �Fracture �Propagation

Plumose Structure Slickensides Crystal Growth

Sense of Shear

Figure 27

Relationships between

fractures and bed thickness

in sandstones and

limestones

Figure 28

Features on fracture

surfaces: Left) plumose

structure indicating brittle,

extensional fracture;

Centre) slickensides

indicating shear; Right)

mineralised surface

indicating an open fracture



Occurrence of fracturesFractures represent the distortion (e.g. a state of strain) that accumulates in rocksundergoing cataclastic deformation. Therefore, they can occur anywhere thatconditions favour cataclastic deformation mechanisms. They often occur as subsidi-ary elements associated with larger structural features, such as faults (Fig. 29). Theycan also occur in a distributed fashion throughout a body of rock; this case wouldsuggest a style of deformation that is not localised.

Synthetic Fractures

Antithetic Fractures

There are important cautions concerning the interpretation of fractures seen in rockoutcrops. Some of the visible fractures may have been produced by processes that areactive at the earth’s surface (e.g. thermal distortions caused by heating/coolingcycles). Others may represent the breakage of rocks related to the relief of stress asthe rocks have been brought to the surface (e.g. by erosional removal of overlyingrocks). Still others may represent ‘true’ deformation caused during the uplift (thismight happen in tectonically-active regions). Of course, some of the fractures that areobserved in outcrops were created by tectonic processes, and these are the ones thatare representative of the subsurface distribution that we are interested in.

Fractures are also important in rock successions that have been folded, or flexed (Fig.30). In folded rocks, fracture density (the inverse of fracture spacing) is sometimesthought to be related to the curvature of the rock layers. The highest fracture densityseems to occur where there is maximum curvature of the rock layers (such as in foldhinges), and this relationship is often exploited to assist in targeting fracture produc-tion in reservoirs.

2.5 Strain FabricsAs noted above, fractures are one expression of strain. There are other mechanismsby which rocks can change their shape (see Appendix), and these produce fabrics inthe rock. ‘Fabric’ refers to planar or linear alignments of minerals or discontinuities

Figure 29

Fractures associated with a

normal fault

Figure 30

Fractures associated with

folding. Note the different

fracture patterns associated

with the inner and outer

surfaces of the layer in the

crestal region

24

(so fractures are a ‘fabric’). Non-fracture fabrics of interest in the petroleum contextare usually found in muddy and shaley rocks. Because these rock types can undergosubstantial flow, there is potential for the platy minerals to rotate and become aligned.Such fabrics can give information concerning the structural history, and especiallyabout the physical conditions that may have existed (pressures and temperatures).These fabrics can also impact drilling - primarily because they are not ‘expected’ - bydeflecting the path of the bit in the same way that normal bedding lithological contrasts do.

2.6 DiapirsThe Earth’s gravitational attraction is one of the most important factors in creating theloads that cause deformation. The mass of the rocks that comprise basins, when actedon by the gravitational acceleration, produces large forces. Variations in density canlead to significant force anomalies, and these, in turn, can produce deformationalresponses.

An important type of structural feature is the diapir (Fig. 31). A diapir is a body of‘flowable’ rock that migrates upwards due to its lower density (compared to thesurrounding rocks). Rock salt commonly forms diapirs, but under-consolidatedmudstones also can become diapiric. (Large granitic intrusions that rise into themiddle crust are also diapiric - while molten, their density is less than that of thecountry rock around them.)

Diapir

Bending of Layers Above

Truncation ofLayers atDepth

Salt and shale diapirs are commonly treated as ‘closed systems’, meaning that weassume that there is no loss or gain of material. Thus, when the diapir rises, it leavesa virtual void somewhere below. Such voids do not, of course, actually exist. Instead,the surrounding rocks subside into the space vacated by the upward movement of thediapir. Because such a process can be long-lived (perhaps a hundred million years),and because sedimentation occurs during this time period, and because the process ofsagging and movement can cause serious distortions of the rock succession (includingboth normal and reverse faulting), diapirs can be structurally very complex. However,diapirs are often good for trapping hydrocarbons. This is because the diapir itself oftenserves as a seal, so that rock layers that are depositionally-truncated against the diapir,or structurally-truncated by its movement, can serve as reservoirs.

Figure 31

Diapir that pierces lower

layers and flexes upper

layers



3. RELATIONSHIP OF STRUCTURES TO GEOLOGICAL EVENTS

3.1 Tectonics and SedimentationSome faults can be active during the deposition of sediments. A common situationoccurs at the margins of rapidly-subsiding basins: the young rocks are weak, and“slide” towards the deeper portion of the basin (Fig. 32). As this motion occurs, therocks are stretched, and large, listric normal faults are produced. Because one side ofthe fault subsides more rapidly than the other side, the rock layers (each representinga time increment) that are deposited are of different thicknesses in the footwall andhangingwall blocks. Such a fault is called a growth fault (the sediment “grows” inthickness across it).

AA'

A'

BB'A

In basins (i.e. where sediments are depositing and accumulating), structures are oftenvery important in terms of localising where deposition takes place. For example, afault at depth might move, producing a sea-floor bathymetric expression (e.g. a low-to-high change in sea-floor elevation). Something similar might happen in a rivervalley in an area of active tectonism. Currents that are carrying sediment wouldpreferentially flow into the lower places, and their sediment load would tend to bedeposited there. Such a sequence of events can be responsible for causing alithological change in the resulting rock sequences: e.g. a lateral change fromsandstone to shale.

The tops of tilted fault blocks, or the crests of folds that are caused by faults beneath,can also have an impact on the distribution of the sediment that is deposited. In a basin,such crests can actually extend above the water, and be subject to erosion. The debristhat is produced during erosion may be deposited locally, leading to anomalous coarsesediments amongst finer-grained materials. If the tops of the structures are notexposed, but instead are at an elevation to cause them to be covered by only shallowwaters, there may be little or no deposition on their crests. If the waters are clear andwarm, reefs might form on the tops of such blocks. Such spatio-temporal variationsin lithology are quite common in sedimentary rocks deposited in tectonically-activebasins.

Figure 32

Syn-depositional faulting

can lead to different rock

thicknesses in the footwall

(eg, layer A) and

hangingwall (layer A’)

blocks

26

In a very broad sense, tectonism (formation of large-scale structures, such as mountainranges and adjacent oceans) usually produces places of uplift and places of subsid-ence. If the uplifted region becomes eroded, the resulting debris tends to betransported to the subsiding area. Thus there is an empirical relationship betweenlarge-scale deformation and large-scale sedimentary bodies. As an example, theNorth Sea basin has an abundance of sand-rich sediments that characterise the Triassicand Jurassic rocks. This clastic material is directly related to the rift-tectonics thatinitiated this basin. Later, mud-rich sediments characterise the younger, post-riftdeposition within this basin (but there are additional, limited pulses of sand related toyounger tectonic events).

3.2 UnconformitiesIn the preceding section, we noted that tectonism can cause rocks to be uplifted andtherefore subject to erosion. If allowed to continue, erosion produces horizontalsurfaces (mountains become plains). This process will truncate rock layers that are tilted.

Later in time (perhaps very much later), the horizontal surface, and the rocks that liebeneath it, may once again subside and become a site of deposition. The new rocklayers will be essentially horizontal when deposited, and thus there will be an angulardifference between the younger rock layers and the older, truncated layers that liebelow the old erosion surface (Fig. 33). We use the word unconformity to refer to sucha surface. An unconformity (which is just a boundary) represents missing geologicaltime. The time is “missing” because there is no depositional record of what happened.

Create Structure

Erosion

AngularUnconformity

Deposit New Layer(s)

Because there are other ways that unconformities can occur, we actually call thesituation in the previous paragraph an angular unconformity. Another type ofunconformity is a heterolithic (“different rocks”) unconformity. When an orogenicbelt is uplifted an eroded, the intrusive rocks and metamorphic rocks that characterisethe middle crust become the uppermost rocks in that area. If later deposition occurs,the sedimentary rocks above the old erosion are very different from the crystallinerocks below. This situation is found within most basins, where sediments lie above“basement”.

Figure 33

Development of an

unconformity



3.3 Intrusions (Dykes / Sills), Hydrofracture, VeinsMolten rock can be injected into pre-existing rocks, forming an intrusion. Here, weare primarily concerned with intrusions that occur in thin sheets (Fig. 34). These arecalled dykes (when they cut across the layering; because layering is usually more-or-less horizontal, dykes are sub-vertical) or sills (which are more-or-less parallel withthe layering; sills usually have shallow dips). Dykes and sills may have thicknessesranging from a few 10s of cm to kilometres. Their areal extent can be anywhere froma few square metres to many thousands of square kilometres.

��yyyyyyyyyy��yyyyyyyyyy��yyyyyyyyyy

��yyyyyyyyyy��yyyy

Dyke Feeder Dyke

Sill

�yLayers Lifted Up

These features are of interest because of their mode of formation. They do not fill apre-existing void; instead, they make space for themselves. The pressure associatedwith the molten rock overcomes the least compressive principal stress (σ

3), and the

sharp tip of the intrusion serves to concentrate stresses. In combination, these twoeffects cause a crack (Mode I) to open in the host rocks, and the molten rock flowsforward into the space. Dykes and sills that are very thick are the product of continuedinjection, and the subsequent widening of the space. Because these injection featuresrepresent new “external” material, they result in an increase in the bulk dimension ofthe rock mass in the direction that is perpendicular to the intrusion. Sometimes, thereare “swarms” of dykes - many, many individual dykes that have similar strike - thatare associated with major crustal extension.

The mechanical process of intrusion is essentially identical with a fracturing processcalled hydrofracture. Natural hydrofractures may exist, but the term is usuallyapplied to features deliberately created to stimulate a well. In this situation, high-pressure fluids are pumped into a rock layer, with the fluid pressure being sufficientlyhigh to cause fracture of the rocks. Usually, the fluid is pumped along with solids thatserve as proppants (glass beads or grains of sand). This is necessary because, unlikea dyke, where the intruding fluid continues to occupy the crack, the aqueous fluids ofa hydrofracture will dissipate into the rock, allowing the newly-formed crack to close.In order to make the new crack a suitable conduit for increasing the flow ofhydrocarbons, the proppant is a critical component.

Figure 34

A dyke crosses the layers. A

sill is intruded parallel to

the layering (but has a

feeder dyke)

28

Elevated fluid pressure can create an effective state of stress that is capable of causingrocks to fracture - even though the framework (“solid”) stress is below the levelnecessary to cause deformation. In addition, fluid pressure affects the stability of pre-existing discontinuities - such as fractures that are already present in the rock mass.(see Appendix) This knowledge is exploited in drilling, where the leak-off test is usedto estimate the magnitude of σ

3. Data obtained from such tests can be used to establish

a “fracture gradient”, which is the plot of wellbore fluid pressures at which fluid losswill occur due to opening of the void spaces (Fig. 35). As a VERY rough rule of thumb,the fracture gradient is approximately 0.8 of the lithostatic gradient (which is taken tobe 1.00 psi/ft).

Pressure

Dep

th

Lithostatic Pressure

Hydrostatic Pressure

Fluid Pressure(showing overpressurebelow seal)

Leak-Off Pressure

"Seal"

Tectonic fractures (frequently, the extension fractures, or joints) may be filled orpartly-filled with minerals (e.g., calcite or quartz). Such features are called veins.Veins are significant because they indicate that the rock mass was dilated (its volumewas expanded).

There are two ways that veins can form. In the first, a fracture is opened slightly.Fluids then circulate through the open crack, leaving behind minerals that were insolution. In the second type, there is very little opening at any time. Circulating fluidsrepeatedly deposit minerals that fill the small opening, and then these minerals arecracked to re-form another small opening, with continued fluid movement. Theminerals that are deposited form fibres that are elongated in the direction of crackopening. The process is known as “crack-seal”. It is more common in high-pressure/high-temperature environments (metamorphism). The crack-seal process gives arecord of actual motion, which is usually not known in the more common types of veins.

3.4 InversionIn basins, structures are important in terms of creating space within which newsedimentation can occur. In many basins, listric normal faults, and associated growthfaulting, are very common. If at a later time, the basin is the location of a different typeof tectonic event (such as shortening), it is possible that the previous growth faults may

Figure 35

Hydrostatic and lithostatic

profiles with an

overpressure zone and leak-

off data.



become re-activated as thrust faults. There would be a thickened rock succession inthe hangingwall of the new thrusts, as well as, potentially, a change from reverse tonormal motion on the fault surface (depending on the magnitude of the displacements).This process is called “inversion”. It is not certain if the tectonics works in exactlythis fashion, or if there is a different “true” interpretation that merely appears to be theone described here. There is considerable debate about this notion, but it is importantto know what is being implied by the discussions that may occur around you.

4. PRACTICAL STRUCTURAL GEOLOGY

4.1 TiltingMost of the processes that form structures result in tilting of the rocks. (As noted inthe Appendix, rigid-body rotation is a component of deformation.) Surprisingly, thisaspect of deformation is often not emphasised in descriptions of Structural Geology.The surprise factor is because the tilting of rocks is almost essential for migratinghydrocarbons from their source region to any available traps.

Because hydrocarbons are buoyant (relative to aqueous porefluids; refer to theChapter on Petroleum Play), there is potential energy available to cause them to rise.What is needed is a pathway for them to follow. A tilted carrier bed (nominally, a rocklayer that has characteristics similar to reservoir rocks) can provide such a pathway.Tilting is also important for a special type of trap. Stratigraphic traps occur ifhydrocarbons accumulate in a reservoir rock unit at the point where that rock unitinterfingers with other, non-reservoir rocks. The general view is that the hydrocar-bons gained access to the carrier/reservoir rock layer, moved upwards along it (the tiltcomponent), and became trapped due to the termination of that particular lithofacies.

4.2 Trap ShapesThe primary rationale for the study of Structural Geology is that structures form trapsfor hydrocarbons. Each trap is unique, with its own combination of rock types, thestratigraphic arrangement, the geometry of the rock layers, the reservoir properties,and the timing of structural events in relation to the migration of hydrocarbons.

The most important point to extract from this section is that structures producedifferences in elevation (Fig. 36). This difference can be produced by flexure of thelayers (into an anticline), or by faulting. Rotation (tilting) is frequently observed intraps, but traps can be formed where the rock layers remain horizontal. Flexure canbe produced by faulting, by buckling, by diapirism, and by differential compaction ofunderlying rocks (e.g. sand compacts less than shale, so a sand “pod” will produce abump in the overlying rocks following compaction).

30

Flexure Faulting

sand

Elevation Differences Produce Trapping Opportunities

Not all trap-shapes result in petroleum reservoirs. If hydrocarbons have not migratedto the trap, or if the timing of formation of the structure is later than migration, noreservoir is created. In other cases, the failure to accumulate hydrocarbons is aconsequence of the lack of a seal. For example, a seal may be removed by erosion ofthe crest of a fault block. In other cases, deformation may impair the integrity of theseal.

4.3 Fault-Zone PropertiesFaults are often surrounded by a zone of deformed rocks. These damage zones maybe a few 10s of cm wide, or they may involve a hundred metres or more of thehangingwall and footwall rocks. Fault zones can be characterised by permeabilityimpairment, or enhancement, or both. Often fault zones are sites for fluid flow in thesubsurface, or they may perhaps be the boundaries of regions of fluid flow. As a result,over geological timescales, there are many opportunities for the permeability of faultsto become further enhanced or impaired through diagenetic dissolution or precipita-tion events.

As noted in the Appendix, many rocks yield by dilatant fracturing. The fracturesproduced this way tend to be conduits for fluid flow. If a fault damage zone consistsof fractures of this type, there can be a permeability enhancement - generally alignedwith the orientation of the fault. Other rocks, or perhaps the same rocks that aredeformed under different conditions, can yield by compactant modes of failure. Forexample, in porous sandstones, shear fractures and small faults are characterised byarrays of small-scale (up to about a mm in width) slip surfaces that have offsets of afew millimetres to perhaps a few centimetres. These features can be observed in coreas well as outcrop, and they have been produced in the laboratory. They are calledcataclastic slip bands (CSBs), or granulation seams, or microfaults. In the usual case,CSBs are characterised by breakage of the original detrital grains (and the limitedcement that is present in the rock), so that there is a grain-size reduction within the zoneleading to a permeability decrease (Fig. 37).

Figure 36

A range of trap shapes

produced by structures



��yy�y��

yyyyyy

��yy��yyyy��yy�y

��

yyyyyyyyy��yyyyyyyy

��

yyyyyyyy

��

yyyyyyyyyyyyyyyy

��

yyy

��

yyy

��

yyyyyy

��yy�y�y��yy�y��yyyy��yyyy

Outcrop Core Thin Section

0.002 m0.1m

0.1m

The granulation process can result in a reduction of permeability by an order ofmagnitude (or more). Preferential cementation may occur at the granulation seam(because of the reactivity of the fine-scale particles, and because of differences influid-flow characteristics), reducing the permeability even further (by several ordersof magnitude). It is important to note that these structural phenomena occur most oftenin clean, porous sandstones (nominally, excellent reservoir rocks) where they canhave a detrimental effect on the reservoir quality. In two-phase flow, the fine porethroats associated with these networks can lead to high residual oil saturations becauseof capillary trapping.

4.4 Fault SealingBecause faults do not have constant displacements, there is a progressive variation inthe footwall and hangingwall rocks that meet at the fault plane. At the edges of a faultsurface (its tips), a rock layer is effectively continuous, because fault displacement iszero. However, as the fault displacement increases (towards the middle of a faultsurface), rocks from different levels, higher and lower in the stratigraphic succession,are juxtaposed due to the fault movement. If some of these rocks are capable of sealinghydrocarbons, while others are reservoirs, it becomes a non-trivial exercise to predictif trapping will occur along the fault. A graphical method to assist this problem isknown as the fault juxtaposition diagram (Fig. 38). This diagram projects bothfootwall and hangingwall rocks onto a 2D view of the fault plane. Visual inspectionallows one to estimate the extent of sealing and communication across the faultsurface.

Shaded = Reservoir, White = Seal (Shale)Grey-on-White Overlaps = Sand / Shale,Grey-on-Grey Overlaps = Sand / Sand

Distance Along Fault

Fault tip

Fau

lt T

hrow

= O

ffset

of L

ayer

s

Figure 37

Granulation seams, or

cataclastic slip bands, in

clean sandstones

Figure 38

Example of fault

juxtaposition diagram

32

A related topic concerns shale smearing along faults. The principal idea is that rocksuccessions with higher percentages of shale are likely to produce clay smears alongthe fault plane as a consequence of fault motion. This is because shales are often veryweak and are easily deformed (Fig. 39). Techniques now exist to calculate a parametercalled the shale gouge ratio (SGR). This number represents the proportion of shale thathas moved past any point, expressed as a percentage of the total displacement on thefault. Rock sequences with high shale fractions, and larger fault displacement,produce higher shale gouge ratios. The numerical value is locally calibrated againstexperience: some reservoir rocks contain hydrocarbons, while others have only water,and the value of SGR that separates these cases is used to predict other trappingsituations in the local area.

��

yyyyyyyyyyyyyyyyyyyy

4.5 OverpressureA condition of overpressure is said to occur when the fluid pressure at depth is higherthan the “expected”, normal hydrostatic pressure (e.g. Fig. 35). There are a numberof mechanisms that can cause this situation to occur:

• Dis-equilibrium compaction — porewater is unable to escape from lowpermeability (usually shale) rocks during subsidence due to burial that is too fast

• Source rocks generating hydrocarbons• Uplift of sealed units that have been buried (and which acquired high pressures there)• Oil or gas columns• High pressure generated elsewhere is communicated to the site

For rocks that are undergoing compaction, overpressure stops the compaction(deformation ceases, or slows). For consolidated rocks, overpressure can createfractures, or cause existing fractures to open, or perhaps to slip. Thus, there aresignificant impacts on rock mechanics whenever overpressure occurs. Overpressureis a dynamic phenomenon that can be found in nearly every basin.

High pore pressures cause difficulties in drilling, and if they are encountered withoutproper precautions, a blowout can occur.

4.6 Stress-Sensitive Reservoir BehaviourThe existing (in situ) state of stress in a reservoir may very well be different from thestress state(s) that existed when the reservoir was created, and when it acquired itsfracture suite. Knowledge about the in situ stress state can be derived in several ways.During drilling, the state of stress causes part of the borehole wall to be unstable, and

Figure 39

Shale smear in a sand-

shale sequence



hence to be lost during drilling operations. The resulting oval borehole shape can bemeasured with four-arm caliper tools, and from the direction of the long dimension,the horizontal maximum stress direction can be inferred. Hydrofracturing, and packertests, can provide additional information.

There is an empirical observation that the in situ stress state affects the production ofa reservoir. Fractures, and perhaps pores, are subject to distortion due to the in situstress. Fractures that are parallel to the σ

1 - σ

2 plane are most favoured for being open

(because their opening is resisted by the smallest stress, σ3). This directional character

of flow can be used to infer the stress orientation. Similarly, it is possible to developreservoir management plans that exploit this phenomenon (Fig. 40). The anisotropyof seismic shear waves may also be used to detect the preferential opening of fractures.

Well

FMS FractureOrientations

Directional Permeabilityfrom Interference Tests

0 0.5 1.0 mi

N

C.I. = 200 ft

Another type of stress-sensitive behaviour concerns rocks that are “under-com-pacted”. Loosely-consolidated sandstones are sometimes observed to have very highporosities. When the reservoir is produced, the pore pressure falls, and the under-compacted rocks then proceed to compact. During this process, loose grains of sandcan migrate to the wellbore, where they are produced along with the fluids. Chalkreservoirs can also exhibit such production-induced compaction. The compaction canbe expressed at the earth’s surface by subsidence, which can be a serious problem -for surface production facilities, including platforms, and for any other humanactivities.

4.7 Fractured reservoirsA fractured reservoir is one in which naturally-occurring, dilatant, conductingfractures (not CSBs) have a significant effect on the producing characteristics of thereservoir — usually through increased permeability. Nelson (1992) gives four typesof fractured reservoirs (Fig. 41):

Figure 40

Example of a fractured

reservoir in which the

direction of the “open”

fractures, as evidenced by

interference tests, is used to

infer the σ1 orientation

34

• TYPE I: Fractures provide the essential porosity and permeability to thereservoir. These reservoirs often deplete rapidly and are not very economic.

• TYPE II: Fractures provide essential permeability. Matrix porosity supportsthe fracture flow to maintain performance and provide sufficient reserves.

• TYPE III: Fractures add to the permeability. Fractures enhance the reservoirperformance, significantly improving the otherwise poor-quality reservoir.

• TYPE IV: These are normal matrix reservoirs where fractures may introducesome anisotropy or compartmentalisation. Fractures of some sort are to beexpected in all reservoirs.

100 %Fractures

AllFractures

100 %Fractures

100 %Matrix

AllMatrix

100 %Matrix % of Total Porosity

% o

f Tot

al P

erm

eabi

lity

I

IV

III

II

Effective flow properties in a realistic fractured rock mass depend on the geometry andintersections of fractures belonging to multiple sets. Fracture porosity is usually <1%in reservoirs. Fracture permeability can range from a few mD to several Darcies.There are two useful expressions for estimating fracture permeability and porosity, asa function of fracture aperture and spacing, for sets of parallel fractures:

kf = a3

d. 8.35.109

ϕf = aa+d

. 100

wherekf = fracture permeability (mD)ϕf = fracture porositya = fracture aperture (cm)d = fracture spacing (cm)

Figure 41

Crossplot showing the

relative contributions of

matrix and fractures in of

fractured reservoirs (after

Nelson, 1992)



4.8 Structural Observations and Interpretations from Typical DataSeismic reflection data (see Chapter 5) represent the primary information source onthe structural forms that are present in petroleum basins. After the seismic reflectionsare correlated with specific rock layers, it is possible to determine the strikes and dipsof those layers (in a grid of 2D seismic lines, this process can be error-prone!). Wherethe orientations change, we can expect the rocks to be damaged. Where layers are notcontinuous, we can interpret faults. Where seismic attributes change, we can(possibly) infer a change in fracture intensity, etc. Because faults are often steeply-dipping, they are usually not well imaged on seismic data. Instead, faults areinterpreted as the junction between regions of the reflection section where layers areobserved to be intact. Very good seismic data can be subjected to fault-plane mapping,which identifies faults ‘directly’.

Geological maps and cross sections are also used to identify structural features. Thesedata forms are particularly valuable for determining the structural style of an area.Subsurface maps and cross sections (see Chapter 7) are also used to interpret the style.However, a caution is in order. Many maps and sections are made to address a specificneed. They may not be appropriate for another need. For example, a regional-scalestructural-contour map produced from widely-spaced 2D seismic lines may revealmajor anticlines (exploration stage), but that same map is probably unsuitable fordetermining curvatures, and hence fracture intensities (production stage). Whereverpossible, the history of a map or cross section should be investigated before itsinformation is used.

Natural fractures can be identified in cores by visual inspection (they have to bedistinguished from drilling-induced fractures), on logs (image logs are invaluable), byloss of circulation during drilling, and by anomalous flow rates during production ordrill-stem tests (production logs can identify flowing fracture zones). Fracture zonescan sometimes be identified on seismic records (because fractures can change thephysical characteristics of the rock). However, evaluating fractured reservoirsrequires the integration of sparse, often non-quantitative data from many disciplines.

Faults can be identified on wireline logs by constructing correlation panels, and fromthese, determining missing or repeated portions of the stratigraphic succession. Thiseffort is made considerably more difficult if the rock layers are dipping, or if thewellbore is highly deviated. This is because the true thicknesses are stretched on thewireline logs. Dipmeter logs can be used to produce True Stratigraphic Thicknesscorrections to the log suite, so enabling an improved correlation and identification offaults.

4.9 BalancingThe principle of structural balance is merely a geologically-phrased version of theclassic law of physics: the conservation of mass. To paraphrase: structural balancemeans that the materials that existed before deformation are still present afterwards,although they are re-arranged (and possibly may have left the local area!). In practice,balancing is used to assist in judging the validity of structural interpretations ofdeformed rocks. Since we always lack complete exposure, alternative ways ofinterpreting the missing (unseen) parts of structures can be compared through thisapproach. This is done by attempting to restore the deformation, and then assessing

36

if the pre-deformation geometry looks plausible. A key point in balancing is theidentification of a pin line: that is, a place where there has been no movement.

Although structural balance sounds like a panacea, in practice, there are manypotential pitfalls. These are related to the simplifications of the concept that areroutinely adopted (such as: line-length balancing, area balancing, etc). For ourpurpose, we can accept that balancing has the potential for assisting in developinggeometric interpretations of structural features, and in deriving interpretations of thedevelopment history of those structures. However, the actual techniques requirespecialist training, and are not presented here; Petroleum Engineers should consult thegeological staff to assist in any balancing work.

4.10 Structural familiesThe study of Structural Geology is merely one aspect of efforts to understand thecomplete geological history of an area. In the context of Petroleum Engineering, thathistory includes the operation of the Petroleum System, including the deposition ofsuitable source rocks, seals, and reservoir units, and their deformation into trapgeometries at a suitable time to catch the migration of hydrocarbons. This sectionfocuses on the idea that the geometries and kinematics of deformation often producecharacteristic patterns. In other words, there are common associations of structuraltypes. After a given style is identified (in terms of associations that are “typical”),reasonable predictions may be made of structural forms in locations where informationis sparse.

The idea of repeated associations is sometimes referred to as structural families(sometimes also called structural styles). Although everyone accepts the notion of theexistence of structural families, there is no universally-accepted list of them. Thereason for this is that the definition of a style is often linked with ideas concerning theprocess of formation of structures. For a decade or so, most types or styles are“explained” via some plausible model, but new research seems to uncover difficultieswith these ideas, and new models of formation are then created. Therefore, there isalways a degree of controversy concerning the origin of some structures (dependingon the point in this idea/revision cycle), and it is difficult to tell students “the” answer.The succeeding sections give (some of) the characteristics of the primary groups ofstructural families.

Fold/thrust beltsFold and thrust belts are long (100’s to 1000’s of km), curvi-linear zones ofdeformation within which there is substantial shortening (typically, the belt is onlyhalf of the original width). These belts usually occur in zones where major platesconverge. In a major orogenic system, there may be two sub-parallel thrust belts: oneon each ‘margin’ of the orogenic belt. Some key characteristics of fold/thrust belts are(Fig. 42):

• Asymmetric folds, truncation- and buckle-folds

• Décollement (detachment-style) thrust/reverse faults

• Good strike continuity of structures, moderate dip and depth continuity



• Generally forms an arcuate belt

• Often a narrow(ish) zone of deformation located at a continental margin (or aformer margin)

• Progressive intensification of structural complexity and metamorphism fromthe externides to the internides (foreland to hinterland)

Fold and Thrust Belt

Horizontal reference

Cover

Basement

100's of KmInternal(Hinterland)Faulted, and even re-mobilised, basementRe-folded metamorphosed cover rocks

External(Foreland)

Fault-associated foldingBuckle folding

��yyyyyy

��yyyy

Long-travelled thrust sheets

Nappes

Wrench provincesWrench provinces in continental terrains can be broad (1000’s of km in each direction)regions produced by the lateral movements of major plates. In oceanic regions,wrench zones typically are more narrow (100’s of km), but possibly very long.Because the wrench faults are usually not ‘perfect’, but instead are irregular (jogs,changes in strike, etc), there is a considerable degree of associated deformationnecessary to allow rock masses to move past one another. Some key characteristicsof wrench provinces are (Fig. 43):

Figure 42

Structures typical of fold/

thrust belts

38

• Intermixing of folding and faulting

• Fault sense (normal to reverse) and fold shapes change rapidly

• Variety of trends (refer to “shear zones” in Appendix)

• Often a strong localization of deformation, with intervening regions of little orno damage

• Vertical and horizontal predictability is poor (even though there may be astrong preferred orientation of fault trends)

• Sometimes adopted as a style whenever there is great complexity, even thoughdefinitive evidence of wrench movements may be lacking

• Regional context is very important in identifying this style correctly

• Contains examples of structural forms of all types

Wrench Structures

Releasing bend Restraining bend Dilational jog Restraining jog

Map expression of wrench zone

Cross section of flower structure along wrench fault

Fault tip

Fault tip

Faultcontinuing

Block diagram of local uplift (Restraining bend or jog)

Reverse fault possibly oblique slip

Anticline often on echelon

Erosion =sediment supply

Down dropping via

faulting or flexure

Fault continues

Fault continues

monocline

MountainsSediment supply

Sediment supply

Slide block

Depositionalarea

Coarse debris

Normal faults

Oblique - slip faults

Volcanics

Map of local depocentre (releasing bend or dilational jog)

Away Toward

Basement

Rifting and extensionRifting and extension represent the lateral stretching of the crust, caused by its ownself-weight, when the lateral constraints are removed (sometimes by stretching of themantle underneath caused by a rising plume). Rifting and extension often occur at thesites of previous orogenic shortening (where the crust was thickened in the shorteningphase) as these later spread sideways and return to normal crustal thickness. Some keycharacteristics of rifting and extension are (Fig. 44):

Figure 43

Structures typical of

wrench terranes



• Tilted fault blocks, often with considerable asymmetry

• Syn-tectonic deposition, and possibility of major and/or rapid changes indepositional setting (non-marine to deep marine, and possibly back to non-marine)

• Listric faults, and abundant smaller-scale faults

• Moderate strike continuity, but abrupt terminations (transfer zones)

• Moderate dip continuity, but variable depth predictability (location ofdécollements)

• Volcanics

Transfer fault

List ric normal fault

Antitheticfault

Rollover anticline

Half-Grabentilted left

Half-Grabentilted right

Detachment fault

Transfer fault

"Fan" of bedding dips insyn-tectonic sediments

Faulting/folding ofshallow layers abovefault-block edges

Headwall

Metamorphic rocks riseand warp detachment fault

Core complex Extreme tilting

Extreme Rifting Leading to Detachment Faults and Metamorphic Core Complex

Rifting Structures

Diapirs and growth structuresDiapirs and growth structures often occur together. They are typically found inpassive-margin settings (a passive margin is produced when a rift is successful atseparating a plate). The evaporites that produce the major diapirs are deposited in theearly rift-stage. They become mobilised and affect deposition and deformationthereafter. Major growth faults are often located on the basin-ward flank of majordiapirs. Modern interpretations of the Gulf of Mexico margin suggest that salt diapirscan move laterally through portions of the basin, leaving behind complex ‘welds’where sedimentary sequences of differing ages are juxtaposed across the former siteof the salt tongues. Some key characteristics of this style are (Fig. 45):

Figure 44

Structures typical of rifting

40

• Typically present in passive margins, but can occur elsewhere

• Depositional thickness changes, rollover structures

• Synthetic and antithetic fault sets

• Predictability is proving less good than was thought only a short time ago

• Growth faults often on flanks of diapirs

• Diapirs of both salt and mud activated by density differences

• Toes of major sheets may show contractional structures that are contemporaneouswith headwall extension

0 50 kmCrustal-scale view

Oceanic crustContinental crust

Salt tongue

"weld"

Overpressed shale

Growth fault system

50-100km

Toe of slide sheet

Thrust faultsand buckle folds

SaltSalt

Wedge of passive-margin sedimentswhich contain these other structures

“Drape” folds / Block faultingAn interesting class of structures that is often not separately considered is that of“drape folds” and block faults. In these features, a deep-seated, high-angle fault (inbasement rocks, or in deeper portions of the basin) causes bending of the overlyinglayered rocks at shallow structural levels. The kinematics of this structural style haveled to heated arguments. This is because there is a discrepancy between the line-lengths of the layered rocks that are folded, and the line-lengths of the faulted rocks.

Figure 45


“passive margins”



In other words, these structures do not follow the rule of ‘structural balance’. (Whatis actually going on is that there are significant volume strains that affect line-lengthsand areas of cross sections. Such strains are not, typically, included in balancingmethods. In addition, there is a considerable lateral motion associated with flow ofductile rocks.) The transition from fault to fold is accomplished via the ‘flowable’ unitwithin the sedimentary succession.

Structures of this type are actually quite common. They characterise the style of intra-cratonic basins and many shallow-shelf and platform regions. There is a growingappreciation of this style in rift environments. This style probably occurs in manyother places, but the structures may have been interpreted to be some other structuraltype. Some key characteristics of this style are (Fig. 46):

• Wide range of fault orientations

• Depending on cause of faulting in basement, lateral predictability varies as perabove styles

• Depth predictability is generally good

• Key element is unit of ductile/weak rock between “basement” and layeredrocks of the “cover”

• This style is probably ubiquitous throughout the “stable” regions of continentsas intra-continental basins form and evolve

weak

strong

basement

attenuation of strong layer possible new layers deposited during fault movement

strong

weak

strong

basement

nearly continuous strong layer

weak

strong

basement

stretching and faulting of strong layer

Normal Fault

Vertical Fault

Reverse Fault

Basement / Cover (Sediments) RelationshipsExamplesof Basement Fault-Block Patterns

Figure 46


platform areas

42

5. FURTHER LEARNING

Exercises

1. Collect together sketches and/or photographs of faults from: local outcrops, thegeological data you have around your office, or published examples in the literature.In each case identify the hangingwall, footwall, sense of displacement, and throw. Ifpossible consider the 3-D aspects of the features. Try to classify the type of faults, andexplain their structural settings. You should have between 10 and 20 examples at arange of scales.

2. Collect together sketches and/or photographs of fractures/CSBs from: localoutcrops, the geological data you have around your office, or published examples inthe literature. In each case identify the fracture sets, any fracture assemblages thatmay be present, and seek to determine the directions of the lengthening andshortening produced by the fracturing. You should have between 10 and 20 examplesat a range of scales.

3. Collect together sketches and/or photographs of folds/flexures from: local outcrops,the geological data you have around your office, or published examples in theliterature. In each case identify the shape of the flexure, any associatd fractures orfaults, and seek to determine the tectonic setting. You should have between 10 and20 examples at a range of scales. If the flexures are related to faults, think about thesequence of formation.

4. Read several papers that describe a field. Identify the structural elements that areimportant to create the trap, or that have a significant effect on production. Do all ofthe features seem compatible within interpretated structural setting?

GeoCh4

Documents

structural geology41

structural geologyrelevant

structural features3

role of structural geology

modern structural geology

review of structural

structural history burial

practical structural