INTRODUCTION Today, in mature petroleum provinces, we seldom or never find classical anticlines with simple four-way closure. All those structures have been found long ago. Nowadays, our standard fare consists of structures bounded in part by dip and in part by faults. We must become expert in recognizing such structures, and in deciding whether or not they have potential as hydrocarbon traps.
Conduits and Seals Where a crack occurs in the earth, we usually call it a fracture. Where the rock layers are displaced across the fracture, we usually call it a fault. The importance of the fault, as distinct from the fracture, is that the fault has greater potential to form a trap ( Figure 1 ).
Figure 1
At the time of cracking, the fault is likely to provide an escape for water (and possibly hydrocarbons) contained in the rocks below; it provides a conduit. For the fault to form a trap, however, the fault must provide a seal. An attractive situation arises where a fault is a conduit in its lower reaches, and a seal in its upper reaches. This may happen, for example, if the conduit formed by the crack allows the escape of mineral-rich waters from below, and the conditions of chemistry and temperature cause the minerals to precipitate out of solution as the water rises in the fault. If a seal forms just above a faulted reservoir we have a trap, while if the conduit remains open below the reservoir we have a path by which the reservoir may be charged from deep source rock. The danger in a fault trap, of course, is that the seal may be broken by later reactivation of the fault, so that the hydrocarbons are lost. Sometimes, therefore, we need faults to act as conduits, allowing the charge of a reservoir from a deep source. In other situations, we need faults to act as seals. And our constant enemy is the breach of a fault seal by reactivation of the fault. In Figure 1, we depend on the fault to act as a conduit, to charge the reservoir from the deep source rock. We depend on the offset cap rock at A to provide an updip barrier. But we also depend on a fault seal at B. If this seal does not exist, we have nothing. if it does, the spill point, C, depends on the thickness and offset of the cap rock. In Figure 2, with a thicker cap rock, we do not need the fault seal at B, but it must be present higher in the section, perhaps at D.
Figure 2
If it is not present by E, the trap cannot be charged unless both the unconformity surface and its overlying rocks are impermeable-and then only after the development of this impermeability. If the fault has been reactivated as suggested at F, any charge depends on the subsequent sealing of this breach, and can represent only hydrocarbons generated since the resealing. We see, then, the critical decisions we must make in fault interpretation. Is it possible for the fault trap to be sourced? Does the fault plane seal? Did the seal develop before or after the generation and migration of the hydrocarbons? Where the thickness and offset of the cap rock imply a spill point, is the trapped volume sufficient to warrant drilling? Has the fault been reactivated? Is it likely to have resealed? When? Were hydrocarbons still migrating after this time? Having regard to the burial history, should we expect oil or gas? In general, there are uncertainties in our answers, and fault traps always carry an additional element of risk. Some decision-makers, as a matter of principle, will not drill downthrown fault traps, or traps where faults can be seen extending to the surface. Others point to the large number of proven fields having just these conditions. As always, the drilling decision should be based on a reasoned assessment of potential reward and potential risk.
One other seismic decision is very relevant to the assessment of risk: is there evidence of any mobile rock above the trap, in which any fault displacement could be absorbed? Thus, the presence of thick salt above the unconformity in Figure 2 would virtually remove any risk from fault reactivation (though in Figure 1 it might not, because of the silt layer). Similarly, a thick overpressured shale is often seen to absorb quite major faulting. Where the salt or shale layer is thin, however, the fault may be transmitted; then we face the judgment of whether it is likely to be sealed by flowage of the mobile material into the fault zone. All in all, therefore, we cannot afford to dismiss fault traps, but working with them will often intensify our problems-and occasionally cause us sorrow.
Interpretation, Past & Present In the past, a great deal of seismic fault interpretation has been grossly in error. Some of the reasons for this are understandable. · With the poorer quality of old data, reflection continuity was often interrupted by seismic artifacts-changes in surface conditions, refraction blind spots, incorrect statics. To a person raised among mountains, every such break was a fault. · In the regrettable days when geologists and geophysicists were kept apart, many geophysicists did not even know what the geologists knew about the nature of faulting. · What explorationists know about basin dynamics and fault generation has increased enormously in the last decade. · There is always a temptation to give the seismic data the weight of a measurement, while regarding the geological input as mere generalization. Today, we resist this, and insist on an interpretation that satisfies both the seismic measurement and the geological generalizations. Perhaps the clearest example of this is that we would no longer accept any fault interpretation, however "obvious," if it did not satisfy conservation of mass before and after faulting. One of the cheapest ways to find oil today is to reinterpret the old fault maps over producing oil fields in the light of present geological knowledge, and to discover pockets of oil never tapped by the drill.
Round and Round In modern practice, therefore, we are much concerned with establishing the tectonic regime of the area-and how that regime has changed over time-before we attempt final picks on the faults. Here we put ourselves into an interpretive loop: we cannot pick the faults until we know the tectonic history, but we learn the tectonic history, very largely, from the faults. So no fault can be picked in isolation; we search first for faults that show the tectonic message most clearly, we make regional and local inferences about the history, and we go back into all the faults until a coherent pattern emerges.
2-D versus 3-D Seismic Data
Fault interpretation on 2-D seismic data is much less safe than on 3-D data. One reason for this is that 2-D lines are seldom sufficiently close to give definite fault correlations from line to line. Another is that fault interpretation is jeopardized whenever a 2-D line is not perpendicular to the fault. However, we must still be ready to do the best we can on 2-D data. Fault interpretation on 3-D data has given us new insight into the geometry, distribution and behavior of faulting. Once, where we mapped a few long continuous faults using 2-D data, we now see more numerous short, discontinuous faults separated by some sort of transfer zone. Similarly, mapping around faults has changed. The use of 3-D data has allowed us to see the details of structure along and near the fault planes. We have learned that structural contours at the faults are not necessarily the smooth and flowing artistic lines that so many of us were taught in the past.
General Issues in Interpretation Figure 1 illustrates the fault-picking operation at a simple level.
Figure 1
We join the apparent breaks or steps in the reflections with a geologically plausible fault trace, continuing the pick upwards and downwards until the faulting becomes mere flexure, or disappears. Simple enough. But even in this simple case, we are looking for additional messages. Above the fault we are always vigilant for evidence
of reactivation (possibly in the opposite direction), because such reactivation may have broken a seal. However, we also remember the possibility that mere compaction, particularly in shales, can suggest minor reactivation of a fault where none has occurred. Further, the figure illustrates that either type of movement can produce significant local changes in stratigraphy above a fault zone. Figure 2 goes to an opposite extreme; it shows us that, even with fair record quality, fault interpretation can sometimes be extremely problematical.
Figure 2
In the face of the changing intervals between major reflections, the requirement for geological plausibility becomes the key; whatever picks we make, we must have an explanation for them. Just joining the reflection breaks in some arbitrary manner will not do. Even when the fault interpretation appears simple, we have to satisfy ourselves that other faulting is compatible with that interpretation. Are the other faults of compatible type and the same age (acceptable), or of different type and different age (acceptable), or of different type and the same age (problematical)?
As we have agreed earlier, many of the problems of fault interpretation on 2-D data arise because the lines are not at right angles to the faults. This problem is made worse by the fact that many areas have been subjected to several tectonic episodes, each generating a system of subparallel faults-but at different orientations. Figure 3 is a fairly simple illustration of this, with a system of curved extensional faults crossed by a system of vertical strike-slip faults.
Figure 3
If we lay out a rectangular grid of 2-D lines over such an area, some or all of the lines must be at unfavorable angles. Figure 4 illustrates fault picking at a basin edge.
Figure 4
We can see very readily the likely geological development, and the picking is straightforward. However, the earth is also capable of some exceedingly grotesque contortions, and sometimes our courage is taxed ( Figure 5 ,
Figure 5
Figure 6 , and Figure 7 ).
Figure 7
Figure 6
In some cases, situations that would otherwise tax our courage are eased when we can recognize the presence of mobile rocks such as salt or overpressured shale. Thus, the major salt-induced fault in Figure 8 gives us few problems.
Figure 8
An essential step in fault interpretation is the recognition of episodes of extension, of neutral subsidence, and of compression. Thus, in Figure 9, we recognize an orderly subsidence at the edge of a rift basin, and the repeated reactivation and adjustment of the main basin-bounding fault.
Figure 9
On the other side of the basin, Figure 10 shows the formation of a major left-tilted faulted block as a consequence of extension at rift time, before the filling of the basin; such blocks, of course, may represent attractive petroleum targets.
Figure 10
We would also read extension (followed by stable subsidence) into the tilted fault blocks of Figure 11.
Figure 11.
It is equally important that we recognize the evidence for compression. At a regional scale, this may be fairly clear, and confirmed by the mountainous nature of the surface. Occasionally we may even see such thrust situations at prospect scale ( Figure 12 ); the associated rollover can yield a very important structural trap.
Figure 12
A more complex example, involving both compression and extension, is given in Figure 13.
Figure 13.
Many petroleum provinces have been subject to episodes of both extension and compression. When these do not act in the same direction, some sideways movements and adjustments occur, and these create distinctive fault patterns, both in plan and in section. Figure 14 braces us for the difficult decisions that we shall have to make, when we are working in such provinces.
Figure 14
Stress A rock is said to be in a state of stress when a force is applied to it. Earth stress is not simple, but involves total force-per-unit-area for a particular point. What makes the measurement of stress complex is the variety of magnitudes and directions of force that occur within a single volume of rock. Basically, however, we can think of stress as the three-dimensional intensity of force acting at a specified point. At any moment in its total history, a rock has structure. It is also true that for any such moment, it is subject to stresses that tend to alter its properties. In reality, only in deep space would this same rock be relatively free of stress. On earth, the forces that compose stress can be divided into two main typesÑbody forces, which act at every point within the crust and surface forces (also called applied forces), which act only at interfaces between objects and, therefore, are defined only along surfaces. We can better understand how these forces act by considering a single clastic grain buried within the earth's crust. Two types of body forces act on this grain at all timesÑgravity and inertia. We usually ignore the effects of inertia. The force due to
gravity is called the lithostatic pressure. It results from the simple weight of overburden transferred by grain-to-grain contact. If our particular grain becomes involved in deformation, two types of surface forces will also exert their influence on it normal forces, which act perpendicular to the surface of the clast, and shear forces, which act parallel to these same surfaces. Figure 1 (Components of stress on a single grain), shows diagrammatically the breakdown of all these forces that compose the total stress on this single grain.
Figure 1
Since gravity and inertia, by definition, act on every particle at every point in the crust, it is the surface forces that are primarily responsible for the creation of geologic structure and with which we as explorationists are concerned. A rock undergoes deformation when stress causes the displacement of particles within it. This stress can result from body or surface forces, or both. In nature, deformation is almost never simple, since it results from a complex interaction between the chemical and physical properties of a rock mass, its immediate environment, and the rate and intensity at which the displacing forces are applied.
We commonly make use of the concept of a stress field which refers to the distribution of stress acting within a defined body. Such a body can be a single folded layer or a sizable portion of a continent. A stress field is described as homogeneous if the stress at each point is equal. The only situation that normally approximates this near the earth's surface is when stress is almost totally due to lithostatic pressure. "Approximates" is the appropriate term to use here, since there is nearly always some component, however small, of lateral stress in the crust. As a measure of force-per-unit-area, stress is a vector quantity and may be expressed as the sum of various components. If we trade our clastic grain for a hypothetical cube of uniform composition and subject this cube to a progressive deformation ( Figure 2 , Resolution of unidirectional force (F) acting on a cube face into the basic normal (N) and shear (S) components of stress.
Figure 2
Also shown is the idealized physical effect of each component), we see that the total force (F) can be resolved into contributions that act perpendicular to the faces of the cube (N, or normal stress), and parallel to them (S, or shear stress). Normal stresses can be either compressional and tend to compact, or tensional and tend to separate particles within a body. The effect of shear stresses, on the other hand, is to move
particles past one another (as in the bottom right of Figure 3 , Generalized orientation of the principal axes of stress for three basic geologic structures.
Figure 3
In (a) and (b), horizontal compression can be thought of as the actual applied force, while (c) shows the effect of horizontal extension). During deformation, the relative magnitudes of these two stress types will change, such that one type may decrease as the other increases.
Principal Axes of Stress Material behavior science, through its detailed analysis of stress conditions, has derived an important conclusion that has proven very useful in structural geology: for any point in a homogeneous stress field, there exist three mutually orthogonal planes along which all shear stresses vanish and only the components of normal stress exist. These three planes are known as the principal planes of stress, and the axes of their intersection are thus the principal axes of stress. These axes are used to describe what are referred to as the three principal stresses. This ideal triaxial system makes everything simpler, since it allows us to speak in terms of only normal
stresses, i.e., compression, or "squeezing," and extension, or "pulling apart" ( Figure 3 ). In geology, principal stress is usually spoken of in terms of compression, which is taken as positive, and tension as negative. (For materials and engineering science, the opposite is true-i.e., tension is positive, compression is negative.) The three principal axes, or stress directions, are correspondingly written as s1 (maximum principal stress), s2 (intermediate principal stress), and s3 (least, or minimum, principal stress). For our purposes, it is useful to understand four special states of stress: 1. uniaxial stress, where two principal stresses are zero and the other is nonzero 2. biaxial stress, where two principal stresses are nonzero and the other is zero 3. triaxial stress, where all three principal stresses are nonzero
4. pure shear stress, where s1 equals s3 and is nonzero, while s2 is zero. This is actually a special case of biaxial stress. Within the earth's crust, the most common stress situation is triaxial, with s1 > s2> s3> 0. In terms of actual physical effects, al can be thought of as representing the relative compression that acts to deform a rock body by compacting particles into each other, while a represents the relative tension and the stretching of particles away from each other. We say "relative" because these stresses are often defined only in relation to each other. For example, a strong extensional event in the crust will have a al associated with it that is secondary i.e., the compression is passive, related mainly to overburden pressure in this case. The relationship between the three principal axes of stress can be pictured by considering the triaxial examples shown in Figure 3 , which place our hypothetical cube in evolving fold and fault structures. Again, s1 represents the direction of maximum effective compression; s3, extension. At this point, we should not be concerned with the specific angle that the faults make with respect to these directions of stress. Note, however, that these faults and s2 are normal to the plane of the paper. In physical terms, s2 is parallel to the surface of the earth and basically represents the presence of essentially infinite neighboring material. Thus, with these three axes, the conditions of stress at any point or as geologists often apply it at a certain location within a defined plate tectonic regime, can be described and related to actual geologic structures. Though there are no shear stresses acting on the surfaces of the cube in our example, the fact that s1, s2, and s3 differ in magnitude and direction means that shear stress can be resolved within the cube. In fact, the quantity s1 -s3 (s1 minus s3), called the differential stress, is sometimes used as a general indicator of shearing stress. If we were to measure the shearing stresses generated within our cube, we would find that they reach a maximum along planes that are inclined at 45 degrees to s1 and s3 and intersect in the plane of s2 ( Figure 4 ) This represents the ideal case, and (as we will see later on) it helps explain why rocks do not fracture
randomly when stressed to the point of rupture.
Figure 4
Strain Strain is said to exist when particles within a body have undergone displacement. Specifically, strain, usually denoted as 1, compares the change in length, DL, of any line element in a rock body to its original length, L, which increases in some directions and decreases in others. Strain is a measure (qualitative or numerical) of this displacement, which, if permanent, is called deformation. Changes in shape resulting from strain are called distortion; those in volume are described as dilatation, which can be either positive (in expansion) or negative (in contraction or shrinkage). Strain can result from both body and surface forces. Although strain is commonly conceived of as the effect of stress, stress and strain are inseparable during actual deformation. Furthermore, if we make a close comparison between Figure 1 (Computer-generated stress field for a hypothetical
fold in a less viscous matrix.
Figure 1
The short lines represent s 3 axes) and Figure 2 (Computer-generated strain pattern for Figure 1 .
Figure 2
Short lines are drawn perpendicular to the axes of maximum shortening), we can see that the specific relationship between stress direction and resulting strain is complex, even in simple situations. We can see, for example, that the orientations of the short lines in these figures coincide least within the flanks of the folded layer, where shearing stresses are greatest. As we move up or down the flanks, toward the "hinges" of the fold, we notice that the attitude of the strain lines must be explained in terms of a growing combination of normal and shearing stresses. Thus, we would need to understand precisely the mechanisms involved in the transformation of stress into strain before we could accurately predict how these two vector quantities are geometrically related. As a result, we often introduce helpful simplifications into the analysis of strain in nature. The more important of these are based on the assumption that the strain involved in deformation has been relatively homogeneous. Much of the terminology derived for understanding stress has also been applied to strain. For example, our hypothetical cube is said to have suffered homogeneous strain, also called uniform strain, when the strain is the same at all points within it. This means that originally straight lines remain straight after deformation ( Figure 3
Homogeneous).
Figure 3
Thus, for example, a cube becomes a rhombohedron, while a sphere inscribed within it becomes an ellipsoid. In fact, when we describe strain, we often do so with reference to an imaginary object called the strain ellipsoid. In this strain ellipsoid, the principle axes of strain are denoted as l1, l2 and l3 where the maximum lengthening is designated l1 and the maximum shortening as l3. This should also help make clear the basic concept of inhomogeneous strain ( Figure 3 , Inhomogeneous). Inhomogeneous strain is by far the most common in naturally deformed rock. This type of strain involves some amount of rotation in the position of particles, which means that originally straight lines become warped and detailed analysis becomes impracticable. It is, therefore, almost always useful to find some way in which natural deformation can be approximated as homogeneous. The most common approach is to consider geologic structures as the summation of many localized homogeneous strain fields. This method has proved especially helpful in the explanation of secondary rock fabrics, such as mineral alignment and fracturing. Pure and Simple Shear There are a number of basic ways in which deformation by homogeneous strain at constant volume occurs. Those that involve simple flattening and stretching are shown in Figure 4 (Uniform extension),
Figure 4
Figure 5 (Unifrom flattening),
Figure 5
and Figure 6 (Plane strain: Each are basic types of homogeneous strain imposed on a cube of ideally uniform composition.
Figure 6
In each case, the inscribed circle and ellipse represent cross sections through the strain wllipsoid before and after deformation). To understand how we treat natural deformation, however, it is also necessary that we look at the two basic types of shear strain-pure shear and simple shear ( Figure 7 , Hypothetical cross section and diagram to illustrate domains of pure and simple shear in a series of folds that show progressively greater strain.
Figure 7
The shape of the strain ellipse can be the same for either type of shear and cannot be used to derive detailed strain history). Both types help us explain a great many large and small scale features seen in rocks. Pure shear is a form of strain in which no rotation of the strain axes takes place. It is also often referred to as an irrotational deformation. It results from uniform extension in one direction and contraction perpendicular to it ( Figure 8 ).
Figure 8
In simple shear, all particles within a body are displaced in one direction. This is our cube pressed into a rhombohedron again; this time, however, we need to take note of the rotation in the strain ellipsoid. Simple shear can be visualized by imagining the result of placing our cube with its inscribed sphere between the two surfaces of an active fault. The shearing motion created by these two surfaces stretches and flattens the sphere into a strain ellipsoid whose long axis is progressively rotated until it is nearly parallel to the fault plane itself. Displacement within such a body takes place by slippage along closely spaced planes ( Figure 8 ). In actual materials, this can be accomplished in a number of ways for example, by slippage between grains or crystals, or by actual flow at elevated temperatures and pressures. As we shall see, this style of deformation has widespread application to geologic structure.
Rock Strength A substantial amount of literature exists with regard to experimental rock mechanics in the laboratory, the intent of which is to simulate and analyze the process and effects of deformation. Studies are usually performed on cylindrical, core-like samples, which are subjected to compressive or tensile stresses in a chamber whose
pressure and temperature are regulated. From these studies, scientists have determined that deformation generally progresses through three main stages (Nadai, 1950). These stages are defined by the behavior of the deformed material, and are most clearly and simply shown by the use of stress-strain diagrams ( Figure 1 ).
Figure 1
In order, the three stages are as follows: 1. Elastic During this initial stage, the stress-strain relationship is linear. If stress is removed, the body reverts to its original dimensions. No deformation (permanent strain) results; strain is said to be completely reversible ( Figure 1 , segment A). 2. Plastic As stress continues to increase, it will eventually reach some limit beyond which the body suffers permanent strain. This is termed the elastic limit, or yield stress. Beyond this limit, material is said to behave plastically: any increase in stress brings a corresponding increase in deformation ( Figure 1 , segment B). 3. Rupture With continued increase of stress, the body will eventually fracture, rupture or fault ( Figure 1 , point C).
Elastic strain can also end in rupture; a material simply fails without having undergone any measurable deformation. Models of Deformation These three stages through which deformation normally progresses are idealized as the behavior of three hypothetical "bodies" that are subjected to stress, as shown in Figure 2 ,
Figure 2
Figure 3 , and Figure 4 (Stress/strain relationships for several ideal materials: (a) Hookean (elastic) body; (b) St.
Figure 3
Venant (plastic) body; (c) Newtonian (viscous) body).
Figure 4
Each style of behavior is denoted by the name of a well-known mathematician and should be apparent from the stress-strain graphs shown. A Hookean body knows only elastic strain before rupture, and is approximated by a simple elastic spring attached to a fixed body. A St. Venant body, in contrast, shows elastic strain up to a yield stress and then deforms indefinitely by shear strain at that same stress. This type of behavior is approximated by a weight that is pulled across some surface by
an attached spring; the spring stretches elastically up to the point where friction on the table top is overcome and the weight begins to slide. The last example, called a Newtonian body or, sometimes, Newtonian liquid, has no shear strength at all and therefore exhibits no elastic strain. It deforms by what is called viscous strain. In Figure 4 , this type of body is represented as a porous piston pulled through a fluidfilled cylinder. A Newtonian body, then, will deform indefinitely in response to any shear stress, with the total strain being directly proportional to the amount of elapsed time. Strain Hardening These three types of ideal bodies help describe the components in progressive deformation when it is caused by increasing amounts of stress. We might compare them to the graphs shown in Figure 5 , which describe two generalized stress/strain relationships for known materials.
Figure 5
The principal difference between the ideal
( Figure 2 , Figure 3 , and Figure 4 ) and reality ( Figure 5 ) lies in the behavior known as strain hardening. Such hardening is often the result of complex readjustment, or even recrystallization, of the material suffering strain. We can envision it on one level as being due to the compaction and realignment of particles during progressive deformation. In sandstones, for example, stress concentrations along grain boundaries result in extensive grain fracturing and dissolution. This causes the progressive filling of pore space with grain fragments and recrystallized quartz. All other conditions (e.g., temperature and pressure) being equal, this will increase rock strength-that is, greater and greater amounts of stress become necessary to impose the same increment of deformation ( Figure 5 , upper curve). We expect certain clastic lithologies to respond in this manner. On the other hand, strain hardening can occur up to some ultimate strength, after which the stress necessary to cause a given strain decreases continually ( Figure 5 , lower curve). Lithologies such as salt and gypsum-anhydrite may behave this way under certain conditions.
Brittleness and Ductility On the basis of the differences between elastic and plastic behavior, we are able to characterize the general response of materials as either brittle or ductile. Brittle materials rupture before any significant plastic deformation occurs. Such behavior in rocks is marked by the development of breakage discontinuities along the planes that represent maximum shear strain. These are not necessarily faults or fractures visible to the unaided eye, but may take place between individual grains or within crystal lattices. In contrast, material is described as ductile if it is able to undergo a large amount of plastic deformation before failing. It should not be assumed that brittleness guarantees faulting, or that ductility inevitably leads to folding. In every situation, whether a rock responds in a brittle or ductile manner depends on several parameters-composition, effective confining pressure, temperature, strain rate and anisotropy. It is, in fact, relatively meaningless to speak of the true "strength" of a rock without reference to these parameters. Figure 6 ,
Figure 6
Figure 7 ,
Figure 7
and Figure 8 ,
Figure 8
(A series of stress/strain diagrams showing the effects of confining pressure ( Figure 6 ), temperature ( Figure 7 ), and strain rate ( Figure 8 ), on rock deformation on homogeneous (Solenhofen) limestone. Note for ( Figure 8 ): a strain rate of 10-7 per sec is roughly equivalent to 1 year of time; most geologic deformation occurs at strain rates of 10-14 per sec) indicate how laboratory analysis has shown deformation to vary as a function of pressure (a), temperature (b), and strain rate (c) on homogeneous limestone. These laboratory tests tell us that near the surface, at low temperatures and pressures, rock will tend to act in a more brittle manner. With growing overburden, which increases both pressure and temperature, ductility generally increases. At the same time, however, the element of time is crucial. If small amounts of stress are applied over sufficiently long periods of time, almost any rock will deform plastically; only those lithologies with very low resistance to shear, such as basalt, may not. But given normal rates of tectonic deformation, there is some transitional depth range over which the response of a particular rock type will grade from dominantly brittle to ductile.
Figure 9 is a schematic representation of the spectrum from brittle to ductile behavior in limestone, as determined by laboratory tests.
Figure 9
Such testing applies a uniaxial stress (either compression or tension) to a cylindrical sample that is sealed within a chamber whose temperature and confining pressure can be regulated. Limestone and marble have been favored as samples, since these lithologies are more isotropic than clastic rock types. In addition to pressure, temperature and strain rate, the other factor that determines a rock's response to stress is, in one sense, the most obvious-its composition. Figure 10 is a diagram derived by Handin et al.
Figure 10
(1963) from extensive lab experiments on natural rock. Shown are measured ductilities for four common sedimentary lithologies in a water-saturated condition. One interesting point made by this diagram is that, with only one kilometer of burial, marked differences in ductility already exist between limestone and other lithologies. This is partly the result of pore pressure effects that strongly reduce ultimate strength. It is also, however, directly related to the mineral structure of calcite, which allows for significant intracrystalline gliding, even at moderate pressures and temperatures. The same trend of ductility increase is true for sandstones, and is often aided by pore pressure effects. We might also note that dolomite, while slightly more ductile than quartzite at depths below two kilometers, shows the tightest range of permanent strain before rupture. Due to its mineral structure, dolomite does not deform readily by intracrystalline gliding. What is normally referred to as "rock strength," then, is not a fixed property, but a relative response, determined by a specific set of immediate environmental conditions. Since both engineers and geoscientists make use of the terms compressive strength, tensile strength and shear strength, two major points should
be made in relation to these qualities. First, almost all materials are weaker under tension than compression ( Table 1). And second, it is very often the shearing strength of materials that determines when and how they will fail.
Average crushing strength Sandstone 740 Limestone 960 Quartzite 2020 Granite 1480 Basalt 2500
Tensile strength 10-30 30-60 30-90 30-50 -
Shearing strength 100-200 150-250 150-250 150-300 50-150
Table 1: Measured strength of various rock types at standard temperature and pressure
Principal Stress Directions and Faulting Figure 1 ,
Figure 1
Figure 2 ,
Figure 2
and Figure 3 display the idealized orientation of the principal stress axes during reverse (a), strike-slip (b) and normal (c) faulting. Because faulting and fracturing both represent the brittle rupture of rocks, they are often discussed in relation to similar stress systems.
Figure 3
Figure 4 shows the generalized orientation of actual fractures formed in a experimental triaxial test, in which a block of Solenhofen limestone was shortened by 1 percent at room temperature.
Figure 4
The two categories of fractures that explorationists usually speak of, shear fractures and extension fractures, were generated by this test and are indicated. This is how rock generally ruptures in the laboratory. Note that a broad correspondence exists between ideal faulting and shear fracturing with relation to stress orientation. As we have previously noted, rocks are weakest in shear. Two mutually orthogonal types of extensional ruptures are shown in Figure 4 . These have been related to the stresses generated during loading and unloading, and, in each case, they develop normal to s3. Notice that during unloading, s1 and s3 exchange the orientation that they had during loading. This tells us that faults can conceivably develop at very low angles to s1, and that we should expect the relaxation of tectonic stress to generate late-stage features, especially fractures, in a rock body. More generally, it can be understood from Figure 1 , Figure 2 , and Figure 3 that a region undergoing a high degree of faulting is characterized by numerous local stress fields that change in both orientation and magnitude as diastrophism progresses. The geometry and attitude of particular fault planes, therefore, may not always appear directly related to the stress axes that apply to large-scale structural trends. The stress orientations shown in these figures do not, by themselves, always explain
the specific angles at which faults develop or the shape of fault planes. For example, many fault planes are curved and cannot be explained simply in terms of one stress axis alone we need several at varying depths. At smaller and smaller levels of scale, many faults reveal increasingly complex components of displacement, usually involving shear. couples. Figure 5 shows the detailed resolution of shear fractures in an evolving monocline. The fractures shown are those which immediately precede propagation of a high-angle reverse fault from basement into the overlying strata.
Figure 5
While in their basic geometry monoclines represent one of the simpler geologic structures, the details of the deformation associated with them are not simple at all. We should expect, then, that rocks in the vicinity of a fault or fault zone will be complexly sheared for some distance on either side of the fault plane itself. This distance may be measurable in centimeters to kilometers, depending on the lithologies, the type of fault, and the amount and environment of displacement.
General Terminology
Much of the basic nomenclature relating to faults was derived from coal mining in the British Midlands during the late eighteenth and early nineteenth centuries. In fact, the word "fault" itself was originally used by miners to describe the sudden, unexpected, and vexatious termination of a coal seam. Thus, "fault" then carried much of its vernacular sense-some sort of mistake had been made. Early geologists like Murchison and Lyell, however, were quick to realize that a fault was, in reality, a fracture where displacement had occurred, and that simple geometric methods could be used to predict where a particular seam might be found again, either above or below. The bounding surface of a fault presented a "wall" to the disgruntled miner, who was normally forced to continue his heading a short way into solid rock and then start a new shaft in order to relocate the seam. The wall of the fault plane was almost always inclined, which meant that the miner could hang his lamp from the rocks on one side above the fault and rest his foot on those on the other side below it ( Figure 1 ).
Figure 1
Thus, the terms hanging wall and footwall then, as now, simply label the two sides of the fault and imply nothing about displacement. Due to its more common occurrence in the coal-mining area of Britain, a fault was normal to the miner's experience when
it was inclined toward the hanging wall. A coal seam that ended against such a normal fault could be found again if the miner continued his tunnel a short distance in the same direction and then sunk a shaft downward. At times, the reverse was true, and the corresponding fault was designated as a reverse fault. These terms remain in use today. By modern definition, rocks are said to be faulted when they have suffered observable displacement along a plane or interval of rupture. As noted previously, such rupture occurs mainly by shear. The fault plane can be relatively simple ( Figure 2)
Figure 2
or it may consist of a large number of individual offset surfaces and thus be more accurately described as a fault zone ( Figure 3 ).
Figure 3
(In the true mathematical sense, the term "plane" is incorrectly used, for rarely does a fault exist without horizontal and vertical kinks, bends and changes of direction. Because of this, some of our colleagues protest the use of "fault plane" and instead support the term fault surface. Although technically incorrect, fault plane is more commonly used, so that is the term we will use in this text.) Less frequently, rocks may be displaced by a form of shearing that causes loss of internal cohesion but not actual rupture of lithological layers ( Figure 4 ). We use the term shear zone to label fault zones in which the individual planes of displacement are extremely closely spaced.
Figure 4
The great majority of faults, however, more closely approximate planes of slip along which shearing has taken place as a result of movement. Faults are generally classified on the basis of their relative sense of displacement. Yet, as for most of geology, certain settings have encouraged the development and use of more specific terminology, often related to proposed mechanisms. Each major type of fault discussed immediately below is given a more specific nomenclature when discussed in relation to a particular structural style. Displacement varies along a fault, being greatest near the middle of the fault and decreasing to zero at both ends. If we contour the amount of displacement, the resulting map describes a displacement ellipse. We can construct displacement ellipses in both plan view ( Figure 5 ) and along the fault trace.
Figure 5
Net slip along a fault is measured by a vector that traces the displacement between originally adjacent points ( Figure 6 , Basic terminology for fault offset, showing strike-slip (ss) and dip-slip (ns) ).
Figure 6
It is most often resolved into dip-slip (measured in the dip direction) and strike-slip (measured in the strike direction) components. Faults in which one or the other of these components is dominant are correspondingly named dip-slip faults and strikeslip faults. For nonvertical dip-slip faults, we find it useful to again divide the displacement into vertical and horizontal components ( Figure 7 , Frequently used terminology for faults in cross section: (a) definition of throw, heave, and hade.
Figure 7
(Note: the latter two terms are more common in mining geology); and (b) definition of normal stratigraphic separation). From coal mining terminology, the former is sometimes called the throw, the latter, the heave, and the angle between the fault plane and the vertical, the hade. We frequently refer to the vertical and horizontal separation across a dip-slip fault, while the inclination of the fault plane is simply said to be either high-angle or low-angle with reference to a horizontal datum. Also, we often use log data to determine the normal stratigraphic separation, as shown in Figure 7. When "Throw" is Not Throw The definitions noted in Figure 7 are fine when the beds are horizontal. However, when they are dipping, we must introduce an additional term, called vertical separation, and distinguish it from throw. Figure 8 shows a block diagram of a normal fault in which the formations are dipping from right to left.
Figure 8
In this figure, we can see that when we project dipping beds across the fault plane, we can measure the displacement in two ways-we can project the bed across the fault perpendicularly to the vertical plane, or we can project the bed in the same dip angle across the fault. The former is the throw, the latter is defined as the vertical separation. There are only two cases where the vertical separation equals the fault throw. The first is the case as shown in Figure 7 in which the stratigraphic units are horizontali.e., zero dip. The second is when the fault is vertical. When we measure the fault "throw" using well logs, we are actually measuring the vertical separation. However, when we measure the fault "throw" using a seismic section, we are usually measuring the throw, although we are actually measuring the apparent throw. The apparent throw is a measure of the fault displacement as a function of the angles between the direction of the seismic line and the strike of the fault as well as the dip of the beds. Thus, if several seismic lines are oriented at different angles relative to the strike of a fault, each line would yield a different value of the fault displacement. The fault throw," as measured on a seismic line, is the true fault throw
only if the direction of the seismic line is perpendicular to the fault strike and the bedding horizontal. Ignoring this question of relative orientation, we can use seismic data to determine the vertical separation (actually apparent vertical separation) to allow us to integrate seismic measurements with well log measurements. Figure 9 (Determination of vertical separation by projecting dipping reflections across the fault plane), shows that we can do this by projecting the interpreted horizons across the faults.
Figure 9
Back to Displacement Figure 10 (Major types of dip-slip faults.
Figure 10
H and F refer to hanging wall and footwall) shows the four types of dip-slip faults. In most cases, it is assumed that simple shear has acted as the principal strain within the fault plane ( Figure 11 , Diagram illustrating the progressive simple shear within an ideal fault plane ).
Figure 11
We should note here that the fault type may not always be obvious. Where separation cannot be determined or appears to change along strike, determining the nature of the fault can be difficult. In some regions, a diversity of interrelated faulting styles exists in close juxtaposition; differentiating specific structures can prove very difficult. In addition, many intracontinental basins are characterized by near-vertical, highangle faults of small displacement that cannot be easily identified as normal or reverse. Furthermore, multiple episodes of tectonism often affect a single region.
These episodes may involve contrasting stress regimes, and displacement along early faults can be reversed. One of the more important occurrences of this phenomenon involves provinces initially dominated by normal faulting being subjected to compressive stresses as a result of changing plate boundary interactions.
Normal Faults Normal faults can either be planar or listric (concave upward). On seismic sections, planar faults often appear curved, due to the effect of increasing velocity with depth. In general, however, normal faults are more easily identified on seismic sections than other types of faults. This is because of their frequent occurrence in deep, dominantly marine basins characterized by otherwise relatively undeformed sediments. Normal faulting on a regional scale is most often referred to as block faulting, since it results in the creation of horst (high) and graben (trench) structural topography ( Figure 1 ).
Figure 1
Refer to Figure 2 ,
Figure 2
Figure 3 , Figure 4
Figure 4
, and Figure 5
Figure 5
.
Figure 3
The normal faults in these figures are clearly visible. In these seismic sections, it is relatively easy to correlate the reflection horizons across the faults and distinguish the hanging wall from the footwall, as well as the (apparent) throw in time. (The throw in depth may be a little more difficult.)
Thrust Faults A thrust fault has often been defined as a reverse fault dipping less than 45 degrees. However, today we use "thrust" as a generic term to imply near-horizontal, tangential compression and a zone of movement dominated by simple shear strain. In Figure 1 and Figure 2 , we see seismic examples of thrust faulting.
Figure 1
In Figure 1, the effects of compression are documented by the presence of older rocks overlying younger ones.
Figure 2
We noticed the "rollover" of beds caused by the thrusting in Figure 2. Thrust faults can (and do) dip at any angle, and individual thrust planes often show complex, curving geometries. They may be essentially horizontal (along incompetent bedding planes), or flat, for kilometers, then "ramp" to a higher structural level in competent rock and, finally, when reaching another incompetent bed, flatten out again . Thrust faults may curl up at their termini to become nearly vertical or even overturned. In the latter case, thrusts become apparent "normal faults," geometrically speaking. Often, due to complexity, it is very difficult or even impossible to distinguish between underthrusting, in which the lower, relatively undeformed "block" is active, and overthrusting, in which the upper, deformed block actually moves. We may only know that shortening is involved, and this shortening can be as great as tens of kilometers or more for single faults. Rocks that have been transported from their original location (or root zone) are said to be allochthonous ("other earth"), while those that remain in place are called
autochthonous ("same earth"). The allochthon is also variously known as a thrust sheet or plate (not lithospheric), or, in certain circumstances, a nappe. Some thrusts define a listric plane that flattens with depth while others are of a ramp-and-flat geometry. Several or more subsidiary thrusts commonly occur within a single allochthon, either as splays off the sole fault (the deepest and controlling major thrust fault in an area) or as local ruptures in the cores of anticlines. In general, the thrusts in foreland belts are very rarely simple listric planes, but are themselves folded and, at times, truncated by younger thrusts. As discussed by Dahlstrom (1970) and Elliot (1976), the geometry and location of thrust faults in thick sedimentary sequences is largely determined by the distribution of competent and incompetent layers. Four simple rules summarize this influence: 1. Thrusting cuts up-section in the direction of displacement. (This is often called the direction of tectonic transport or, in older nomenclature, the facing direction.) 2. Thrusting tends to parallel bedding in incompetent layers, occurring near contacts with competent units, and to cut obliquely up-section in thicker, more brittle units. 3. The age of major faults is younger in the direction of thrusting. 4. Major thrust faults do not overlap appreciably.
Evolution of Thrusting The overall evolution, therefore, is for thrusting to begin at deeper levels and to progress upward and outward (i.e., away from the root zone) from a major sedimentary basin. Ductility contrast in the stratigraphic section encourages a stairstep evolution of thrusts, which we know as ramping. A relatively flat portion of a thrust plane, particularly where it remains parallel to bedding within a single lithology, is referred to as decollement. If this flattening of the thrust plane occurs at the boundary between the sedimentary section and basement, it is called basal decollement. The progressive stacking of thrust faults may develop in piggyback fashion, where younger thrusts form in the footwall or, alternatively, in overstep fashion, where thrusts become younger toward the root zone. Piggybacking seems prevalent on a regional scale and is by far the more significant progression. At the same time, both piggyback and overstep thrusting occur on a more local level, often as a result of imbrication. Imbricates occur most often in two structural positions of high stress concentrationnear the toe of a major thrust and above ramps in a thrust plane. They dip steeply as they approach the surface, and stack slice after slice of the same stratigraphic section along faults, which sole out into a major thrust plane. Continued movement along this plane after the imbricates have formed will rotate them, so they can become vertical and overturned. As discussed by Dahlstrom (1970), imbrication actually offers a basic model for foreland thrusting-as the scale of a cross section is increased to become more
regional, major thrusts themselves become imbricates of the largest faults (i.e., those with the greatest displacement). These, in turn, can be thought of as subsidiary faults to a basal detachment or decollement plane that marks the structural boundary between basement, usually crystalline metamorphic or plutonic rocks, and sedimentary cover. Thin- versus Thick-Skinned Tectonics The concept of basal decollement is the fundamental structural principle in the hypothesis known as thin-skinned tectonics. This theory is often contrasted with the thick-skinned hypothesis, which postulates no sole fault and, there-fore, direct involvement of basement in each major thrust. The debate between these two schools of thought is a historical one that continues today. On the basis of drilling and seismic data, both of which have proved the flattening of thrusts at depth in foreland areas, most explorationists now favor the thin-skinned hypothesis for at least the more medial and distal portions of thrust belts. However, toward the metamorphic core of many such belts, basement rocks are known to be heavily involved in thrusting. This involvement can be very extensive, as in the case of the Himalayas, and may, in fact, control the overall style and evolution of thrusting. Yet some recent studies based on deep-reflection seismic profiles (Cook et al., 1979; Cook, 1982) have strongly favored the thin-skinned hypothesis for basement thrusting as well. Thus, the debate has expanded to focus on two major questions whether thick-skinned faulting occurs at all in foreland belts, and, if it does, what is the nature of the transition between it and the decollement tectonics that characterize the sedimentary cover. The controversy represents one of the major areas of research in contemporary structural geology.
Strike-Slip Faults Strike-slip faults can be simply classified as left-lateral (also called sinistral) or rightlateral (also called dextral) on the basis of the displacement sense ( Figure 1 ).
Figure 1
When facing the fault, the direction in which the opposite side appears to have moved indicates this sense. In cross section, the symbols "A" and "T" are used to mark which side of the fault has been displaced away from and which side toward the observer. The San Andreas ( Figure 2 ) is one of the best examples of an active right-lateral strike-slip fault.
Figure 2
Others include the Alpine Fault of New Zealand and the Atacama system of Chile. Left-lateral faults of this scale are well-represented by the Philippine Fault ( Figure 3 ). These large strike-slip faults are often called wrench faults in general, and transcurrent faults more specifically if they cut across regional structural trends.
Figure 3
Transform faults are strike-slip faults that connect convergent and divergent plate boundaries. Basically, they serve to "transform" the interaction between plates into strike-slip motion. The motion along these faults often includes components of compression or tension. Transform plate boundaries, therefore, are the links that unify the world's spreading centers, subduction zones and collision zones into a single mosaic of movement. In Figure 4, Figure 5
Figure 5
, and Figure 6 , we see the variety of structural types associated with large strikeslip faults.
Figure 4
Later in this text, when we discuss strike-slip faults and their relationship to structural styles, we will see several examples of these structures in seismic section.
Figure 6
Simple Fault Traps More often than not, structural traps containing hydrocarbons result from or are associated with some form of faulting. Although there are numerous, and sometimes unique, types of fault traps, most generally have an appearance similar to one of the simple models shown in Figure 1 and
Figure 2.
Figure 1
General Terminology Much of the basic nomenclature relating to faults was derived from coal mining in the British Midlands during the late eighteenth and early nineteenth centuries. In fact, the word "fault" itself was originally used by miners to describe the sudden, unexpected, and vexatious termination of a coal seam. Thus, "fault" then carried much of its vernacular sense-some sort of mistake had been made. Early geologists like Murchison and Lyell, however, were quick to realize that a fault was, in reality, a fracture where displacement had occurred, and that simple geometric methods could be used to predict where a particular seam might be found again, either above or below. The bounding surface of a fault presented a "wall" to the disgruntled miner, who was normally forced to continue his heading a short way into solid rock and then start a new shaft in order to relocate the seam. The wall of the fault plane was almost always inclined, which meant that the miner could hang his lamp from the rocks on one side above the fault and rest his foot on those on the other side below it ( Figure 1 ).
Figure 1
Thus, the terms hanging wall and footwall then, as now, simply label the two sides of the fault and imply nothing about displacement. Due to its more common occurrence in the coal-mining area of Britain, a fault was normal to the miner's experience when it was inclined toward the hanging wall. A coal seam that ended against such a normal fault could be found again if the miner continued his tunnel a short distance in the same direction and then sunk a shaft downward. At times, the reverse was true, and the corresponding fault was designated as a reverse fault. These terms remain in use today. By modern definition, rocks are said to be faulted when they have suffered observable displacement along a plane or interval of rupture. As noted previously, such rupture occurs mainly by shear. The fault plane can be relatively simple ( Figure 2)
Figure 2
or it may consist of a large number of individual offset surfaces and thus be more accurately described as a fault zone ( Figure 3 ).
Figure 3
(In the true mathematical sense, the term "plane" is incorrectly used, for rarely does a fault exist without horizontal and vertical kinks, bends and changes of direction. Because of this, some of our colleagues protest the use of "fault plane" and instead support the term fault surface. Although technically incorrect, fault plane is more commonly used, so that is the term we will use in this text.) Less frequently, rocks may be displaced by a form of shearing that causes loss of internal cohesion but not actual rupture of lithological layers ( Figure 4 ). We use the term shear zone to label fault zones in which the individual planes of displacement are extremely closely spaced.
Figure 4
The great majority of faults, however, more closely approximate planes of slip along which shearing has taken place as a result of movement. Faults are generally classified on the basis of their relative sense of displacement. Yet, as for most of geology, certain settings have encouraged the development and use of more specific terminology, often related to proposed mechanisms. Each major type of fault discussed immediately below is given a more specific nomenclature when discussed in relation to a particular structural style. Displacement varies along a fault, being greatest near the middle of the fault and decreasing to zero at both ends. If we contour the amount of displacement, the resulting map describes a displacement ellipse. We can construct displacement ellipses in both plan view ( Figure 5 ) and along the fault trace.
Figure 5
Net slip along a fault is measured by a vector that traces the displacement between originally adjacent points ( Figure 6 , Basic terminology for fault offset, showing strike-slip (ss) and dip-slip (ns) ).
Figure 6
It is most often resolved into dip-slip (measured in the dip direction) and strike-slip (measured in the strike direction) components. Faults in which one or the other of these components is dominant are correspondingly named dip-slip faults and strikeslip faults. For nonvertical dip-slip faults, we find it useful to again divide the displacement into vertical and horizontal components ( Figure 7 , Frequently used terminology for faults in cross section: (a) definition of throw, heave, and hade.
Figure 7
(Note: the latter two terms are more common in mining geology); and (b) definition of normal stratigraphic separation). From coal mining terminology, the former is sometimes called the throw, the latter, the heave, and the angle between the fault plane and the vertical, the hade. We frequently refer to the vertical and horizontal separation across a dip-slip fault, while the inclination of the fault plane is simply said to be either high-angle or low-angle with reference to a horizontal datum. Also, we often use log data to determine the normal stratigraphic separation, as shown in Figure 7. When "Throw" is Not Throw The definitions noted in Figure 7 are fine when the beds are horizontal. However, when they are dipping, we must introduce an additional term, called vertical separation, and distinguish it from throw. Figure 8 shows a block diagram of a normal fault in which the formations are dipping from right to left.
Figure 8
In this figure, we can see that when we project dipping beds across the fault plane, we can measure the displacement in two ways-we can project the bed across the fault perpendicularly to the vertical plane, or we can project the bed in the same dip angle across the fault. The former is the throw, the latter is defined as the vertical separation. There are only two cases where the vertical separation equals the fault throw. The first is the case as shown in Figure 7 in which the stratigraphic units are horizontali.e., zero dip. The second is when the fault is vertical. When we measure the fault "throw" using well logs, we are actually measuring the vertical separation. However, when we measure the fault "throw" using a seismic section, we are usually measuring the throw, although we are actually measuring the apparent throw. The apparent throw is a measure of the fault displacement as a function of the angles between the direction of the seismic line and the strike of the fault as well as the dip of the beds. Thus, if several seismic lines are oriented at different angles relative to the strike of a fault, each line would yield a different value of the fault displacement. The fault throw," as measured on a seismic line, is the true fault throw
only if the direction of the seismic line is perpendicular to the fault strike and the bedding horizontal. Ignoring this question of relative orientation, we can use seismic data to determine the vertical separation (actually apparent vertical separation) to allow us to integrate seismic measurements with well log measurements. Figure 9 (Determination of vertical separation by projecting dipping reflections across the fault plane), shows that we can do this by projecting the interpreted horizons across the faults.
Figure 9
Back to Displacement Figure 10 (Major types of dip-slip faults.
Figure 10
H and F refer to hanging wall and footwall) shows the four types of dip-slip faults. In most cases, it is assumed that simple shear has acted as the principal strain within the fault plane ( Figure 11 , Diagram illustrating the progressive simple shear within an ideal fault plane ).
Figure 11
We should note here that the fault type may not always be obvious. Where separation cannot be determined or appears to change along strike, determining the nature of the fault can be difficult. In some regions, a diversity of interrelated faulting styles exists in close juxtaposition; differentiating specific structures can prove very difficult. In addition, many intracontinental basins are characterized by near-vertical, highangle faults of small displacement that cannot be easily identified as normal or reverse. Furthermore, multiple episodes of tectonism often affect a single region.
These episodes may involve contrasting stress regimes, and displacement along early faults can be reversed. One of the more important occurrences of this phenomenon involves provinces initially dominated by normal faulting being subjected to compressive stresses as a result of changing plate boundary interactions.
Normal Faults Normal faults can either be planar or listric (concave upward). On seismic sections, planar faults often appear curved, due to the effect of increasing velocity with depth. In general, however, normal faults are more easily identified on seismic sections than other types of faults. This is because of their frequent occurrence in deep, dominantly marine basins characterized by otherwise relatively undeformed sediments. Normal faulting on a regional scale is most often referred to as block faulting, since it results in the creation of horst (high) and graben (trench) structural topography ( Figure 1 ).
Figure 1
Refer to Figure 2 ,
Figure 2
Figure 3 , Figure 4
Figure 4
, and Figure 5
Figure 5
.
Figure 3
The normal faults in these figures are clearly visible. In these seismic sections, it is relatively easy to correlate the reflection horizons across the faults and distinguish the hanging wall from the footwall, as well as the (apparent) throw in time. (The throw in depth may be a little more difficult.)
Thrust Faults A thrust fault has often been defined as a reverse fault dipping less than 45 degrees. However, today we use "thrust" as a generic term to imply near-horizontal, tangential compression and a zone of movement dominated by simple shear strain. In Figure 1 and Figure 2 , we see seismic examples of thrust faulting.
Figure 1
In Figure 1, the effects of compression are documented by the presence of older rocks overlying younger ones.
Figure 2
We noticed the "rollover" of beds caused by the thrusting in Figure 2. Thrust faults can (and do) dip at any angle, and individual thrust planes often show complex, curving geometries. They may be essentially horizontal (along incompetent bedding planes), or flat, for kilometers, then "ramp" to a higher structural level in competent rock and, finally, when reaching another incompetent bed, flatten out again . Thrust faults may curl up at their termini to become nearly vertical or even overturned. In the latter case, thrusts become apparent "normal faults," geometrically speaking. Often, due to complexity, it is very difficult or even impossible to distinguish between underthrusting, in which the lower, relatively undeformed "block" is active, and overthrusting, in which the upper, deformed block actually moves. We may only know that shortening is involved, and this shortening can be as great as tens of kilometers or more for single faults. Rocks that have been transported from their original location (or root zone) are said to be allochthonous ("other earth"), while those that remain in place are called
autochthonous ("same earth"). The allochthon is also variously known as a thrust sheet or plate (not lithospheric), or, in certain circumstances, a nappe. Some thrusts define a listric plane that flattens with depth while others are of a ramp-and-flat geometry. Several or more subsidiary thrusts commonly occur within a single allochthon, either as splays off the sole fault (the deepest and controlling major thrust fault in an area) or as local ruptures in the cores of anticlines. In general, the thrusts in foreland belts are very rarely simple listric planes, but are themselves folded and, at times, truncated by younger thrusts. As discussed by Dahlstrom (1970) and Elliot (1976), the geometry and location of thrust faults in thick sedimentary sequences is largely determined by the distribution of competent and incompetent layers. Four simple rules summarize this influence: 1. Thrusting cuts up-section in the direction of displacement. (This is often called the direction of tectonic transport or, in older nomenclature, the facing direction.) 2. Thrusting tends to parallel bedding in incompetent layers, occurring near contacts with competent units, and to cut obliquely up-section in thicker, more brittle units. 3. The age of major faults is younger in the direction of thrusting. 4. Major thrust faults do not overlap appreciably.
Evolution of Thrusting The overall evolution, therefore, is for thrusting to begin at deeper levels and to progress upward and outward (i.e., away from the root zone) from a major sedimentary basin. Ductility contrast in the stratigraphic section encourages a stairstep evolution of thrusts, which we know as ramping. A relatively flat portion of a thrust plane, particularly where it remains parallel to bedding within a single lithology, is referred to as decollement. If this flattening of the thrust plane occurs at the boundary between the sedimentary section and basement, it is called basal decollement. The progressive stacking of thrust faults may develop in piggyback fashion, where younger thrusts form in the footwall or, alternatively, in overstep fashion, where thrusts become younger toward the root zone. Piggybacking seems prevalent on a regional scale and is by far the more significant progression. At the same time, both piggyback and overstep thrusting occur on a more local level, often as a result of imbrication. Imbricates occur most often in two structural positions of high stress concentrationnear the toe of a major thrust and above ramps in a thrust plane. They dip steeply as they approach the surface, and stack slice after slice of the same stratigraphic section along faults, which sole out into a major thrust plane. Continued movement along this plane after the imbricates have formed will rotate them, so they can become vertical and overturned. As discussed by Dahlstrom (1970), imbrication actually offers a basic model for foreland thrusting-as the scale of a cross section is increased to become more
regional, major thrusts themselves become imbricates of the largest faults (i.e., those with the greatest displacement). These, in turn, can be thought of as subsidiary faults to a basal detachment or decollement plane that marks the structural boundary between basement, usually crystalline metamorphic or plutonic rocks, and sedimentary cover. Thin- versus Thick-Skinned Tectonics The concept of basal decollement is the fundamental structural principle in the hypothesis known as thin-skinned tectonics. This theory is often contrasted with the thick-skinned hypothesis, which postulates no sole fault and, there-fore, direct involvement of basement in each major thrust. The debate between these two schools of thought is a historical one that continues today. On the basis of drilling and seismic data, both of which have proved the flattening of thrusts at depth in foreland areas, most explorationists now favor the thin-skinned hypothesis for at least the more medial and distal portions of thrust belts. However, toward the metamorphic core of many such belts, basement rocks are known to be heavily involved in thrusting. This involvement can be very extensive, as in the case of the Himalayas, and may, in fact, control the overall style and evolution of thrusting. Yet some recent studies based on deep-reflection seismic profiles (Cook et al., 1979; Cook, 1982) have strongly favored the thin-skinned hypothesis for basement thrusting as well. Thus, the debate has expanded to focus on two major questions whether thick-skinned faulting occurs at all in foreland belts, and, if it does, what is the nature of the transition between it and the decollement tectonics that characterize the sedimentary cover. The controversy represents one of the major areas of research in contemporary structural geology.
Strike-Slip Faults Strike-slip faults can be simply classified as left-lateral (also called sinistral) or rightlateral (also called dextral) on the basis of the displacement sense ( Figure 1 ).
Figure 1
When facing the fault, the direction in which the opposite side appears to have moved indicates this sense. In cross section, the symbols "A" and "T" are used to mark which side of the fault has been displaced away from and which side toward the observer. The San Andreas ( Figure 2 ) is one of the best examples of an active right-lateral strike-slip fault.
Figure 2
Others include the Alpine Fault of New Zealand and the Atacama system of Chile. Left-lateral faults of this scale are well-represented by the Philippine Fault ( Figure 3 ). These large strike-slip faults are often called wrench faults in general, and transcurrent faults more specifically if they cut across regional structural trends.
Figure 3
Transform faults are strike-slip faults that connect convergent and divergent plate boundaries. Basically, they serve to "transform" the interaction between plates into strike-slip motion. The motion along these faults often includes components of compression or tension. Transform plate boundaries, therefore, are the links that unify the world's spreading centers, subduction zones and collision zones into a single mosaic of movement. In Figure 4, Figure 5
Figure 5
, and Figure 6 , we see the variety of structural types associated with large strikeslip faults.
Figure 4
Later in this text, when we discuss strike-slip faults and their relationship to structural styles, we will see several examples of these structures in seismic section.
Figure 6
Simple Fault Traps More often than not, structural traps containing hydrocarbons result from or are associated with some form of faulting. Although there are numerous, and sometimes unique, types of fault traps, most generally have an appearance similar to one of the simple models shown in Figure 1 and
Figure 2.
Figure 1
Faults and Diffractions · We need only Huygens' principle to tell us to expect a diffraction from a fault. In Figure 1 (Looking at diffractions in terms of Huygens sources), at point A on a continuous reflector, one Huygens source interacts with those each side of it to leave only the reflected signal at normal incidence.
Figure 1
At B, the presence of the last Huygens source at the edge tells us that a signal must be observed in the direction C, while the absence of further Huygens sources to the right tells us that the "backward" radiation in direction D will not be "canceled." · Exactly the same message is conveyed, though in a different form, by considering the circular reflection zone on the reflector ( Figure 2, Looking at diffractions in terms of the relfection zone).
Figure 2
The reflected signal at surface point P includes contributions from all of the reflection zone R. That at point Q includes contributions from only a semicircular zone S, so that we would expect only half the amplitude. The diffracted signal at point T includes contributions from a zone that gradually decreases in size as the offset increases. · The classical seismic expression of a fault is obtained by summing these contributions. For the assumed model of a terminating reflector, the result is as shown in Figure 3 (Seismic response to a terminating reflactor)-a plane reflection that decreases in amplitude to one-half at the fault, and a hyperbolic diffraction with its apex at the fault.
Figure 3
The diffraction comprises a forward branch of the same polarity as the reflector and a backward branch of the opposite polarity. · The physical necessity for the two branches of opposite polarity is easily shown by bringing together two such models, as in Figure 4.
Figure 4.
In the limit, when the reflector becomes continuous, the diffractions must cancel to zero. · The diffraction amplitudes make sense. Since the reflection melds smoothly into the forward branch, and since the amplitude of the reflection at the fault is one-half that for a continuous reflection, the amplitude of the forward branch near the fault is one-half, and therefore so is that of the backward branch. In this sense, the half-amplitude of the reflection at the fault may be viewed as the whole amplitude minus the backward branch. Therefore, canceling the backward branch with the forward branch (by summing along the diffraction hyperbola, as in migration) not only removes the entire diffraction but restores the reflection amplitude at the fault. · The above model of a terminating reflector is geologically unreal; a reflector cannot just terminate. There must be an additional contribution from the fault face. However, for a vertical fault, it is still true that there must be forward and backward branches of equal amplitude and opposite polarity, to provide cancellation when two blocks are brought together in the manner of Figure 4 , (The physical necessity for backward and forward branches, as the gap between reflector terminations decreases) . · One situation that is close to the classical model is that of a vertically faulted thin horizontal layer ( Figure 5, The differentiated form of the classical response, obtained from a thin layer (here assumed to be hard) ), where the signal from the fault face is very small.
Figure 5
We remember that if the layer is thin, relative to the seismic wavelength, the reflection response approximates to a differentiated version of the incident pulse; then the forward and backward diffractions also approximate to differentiated versions of the classical response. · The response changes if the fault plane is inclined. Although the diffractions remain grossly hyperbolic, the pulse shape and the amplitude of the forward and backward branches may no longer be the same. However, as suggested in Figure 6, (If the fault is inclined the details of the diffractions change,
Figure 6
but bringing the fault blocks together (a) still requires cancellation of the forward branch of one by the backward of the other (b) ), it is still necessary that the forward branch from the left block should cancel the backward branch from the right block (and the reverse) when they are brought together; therefore, these must have the same amplitude and form, but opposite polarity. The reasonableness of this is confirmed when we remember that the tails of the diffractions contain the fault plane reflections, which must also cancel.
· If we complete the fault, as in Figure 7 (The reflection from the inclined fault plane in (a) is found among the tails of the forward branch of the upblock and the backward branch of the downblock (b).
Figure 7
The impression is therefore of stronger diffractions on the downthrown side), there is ordinarily a strong forward branch from the upthrown block and a strong backward branch from the downthrown block, and the lateral offset means that these must approach each other. Again, these tails contain the fault-plane reflection. · The backward branch from the upthrown block and forward branch from the downthrown block ordinarily remain well separated, as in the figure. However, if there is a rapid increase of velocity with depth, the lower diffraction is flatter, and the two could again approach each other. In this case the zone of confluence would contain the reflection from the underside of the fault planeÑobtainable because the rapid increase of velocity with depth turns the downgoing rays upward again.
· The above observations apply to sharp rupture of the reflector at the fault... as occurs with brittle rocks. With more plastic rocks, there is likely to be drag into the fault This may be normal or reverse, and may be modified by reactivation or by isostatic rebound (footwall uplift). For any type and degree of drag, we can synthesize the response using Huygens hyperbolas. · For normal drag into an inclined fault, as suggested in Figure 8 (Likely appearance of a normal fault with normal drag),
Figure 8
a common effect is that only the forward branch from the upthrown block and the backward branch from the downthrown block remain visible ( Figure 8 ). · Similarly, the dominant diffraction from the layers pierced by a salt plug is often the backward branch of reversed polarity ( Figure 9, The termination of a stratal reflector (a) at a salt plug often yeilds (b) an amplitude increase caused by the concave focusing,
Figure 9
and a diffraction dominated by the backward (whose tail contains the reflection from the salt-positive or negative, as the case may be). · The preceding figures illustrate the diffractions obtained when the faulted layers are horizontal. If the faulted layers dip, the diffraction hyperbolas are tangent to the reflection at the point of apparent faulting, and have their maximum amplitude at this point; then the true position of the fault is through the apices of the diffraction hyperbolas ( Figure 10, The situation where the faulted layers dip, and the distinction between apparent, unmigrated and migrated positions of the fault plane ).
Figure 10
· The figure also shows that the apparent position of the fault plane, joining the terminations of the stratal reflections, does not correspond to the unmigrated position of the fault-plane reflection, which is contained within the diffraction tails. The fault pick that we make on the unmigrated section-from diffraction apex to diffraction apex-is the migrated position of the fault. · On an unmigrated section, an inclined fault-plane reflection must always appear to cut across the reflections it truncates; Figure 11 (The fault plane corresponds (approximately) to the apices of the diffractions. The fault-plane reflection is contained within the tails of the diffractions, and therefore must cut across stratal reflections) is a case in point.
Figure 11
· A full hyperbola without any change of polarity at the apex ( Figure 12, The message of a full hyperbola with no change of polarity is a tight syncline-not in itself a fault ) generally indicates a buried focus (often, a tight syncline)-not a fault (though it may also signal a shallow source off the line).
Figure 12
· In principle, diffractions can help us to distinguish between normal and reverse faults, in that the diffraction seen most clearly has the same polarity as the faulted reflection for a normal fault, and the opposite polarity for a reverse fault ( Figure 13 , Signature of an inclined normal fault-plane reflection (for a positive stratal reflection).
Figure 13
(b) The corresponding reverse fault, with its different-polarity diffractions melding into a negative fault-plane reflection.). In practice, we have our usual problems in being sure of pulse polarities. · Diffractions cease to be hyperbolic if there is a lateral change of overburden velocity ( Figure 14 , Lateral change of velocity (introduced, in this figure, by dipping layers) distorts the diffraction form hyperbolic form). (Of course, so does NMO.) The apex of the diffraction curve may no longer correspond to the fault edge.
Figure 14
· Diffractions generally survive the stacking process quite well near their apices, though their tails may be weaker and lower-frequency than would be expected from simple Huygens synthesis.
Perfect and Imperfect Migration The ideal solution to diffractions is perfect migration, which collapses each of them to its corresponding reflection termination, and thereby restores the full amplitude of the reflection at that point. Drag into faults is correctly represented, and flexures not involving actual rupture are distinguished from faults ( Figure 1, Unmigrated) and Figure 2 , Migrated).
Figure 1
Figure 2
However, 2-D migration can be perfect only if the strikes of all stratal reflectors and of all faults are perpendicular to the seismic line. This requires first that the reflector strikes and the fault strikes should be the same (which happens, but not always), and second, that the dip lines were laid out straight and on dip. In this happy situation, our only remaining problem is that the dip lines must be linked into strike lines-for which 2-D migration must be in error. We also know that 2-D migrated sections do not tie at the line intersections, and that we must demigrate the dip lines to achieve this fundamental and powerful check. However, we can work with 2-D migrated sections on such a grid and produce a correct map. But if the strikes are not perpendicular to the lines, that our general approach would be different. Our preference would be to do the reflection picking on the unmigrated data, to proceed to an unmigrated contour map, and then to migrate the map. The problem with this, of course, is that we cannot afford to sacrifice the clarification of the faults provided by section migration where it is valid. Interpretation therefore becomes an iterative process: · We make a first assessment of the fault correlations from line to line, and of the gross reflection contours.
· We thereby identify those faults on which migration is likely to be correct. · We use the migrated sections over these faults to clarify the fault interpretation. · We improve our picking of the unmigrated sections accordingly. · We proceed to an improved unmigrated contour map. · We migrate the map. As a check, we can see whether the migrated map explains the invalid migrations that must be apparent on the migrated sections crossing faults and other features at unfortunate angles. To this end, we need to be able to think out the signatures of incorrect migration. As the reference, the diffraction in Figure 3, from a fault perpendicular to the line, is perfectly collapsed (b) by migration with the correct velocity.
Figure 3
The undermigration in (c) can be caused only by the use of too low a velocity. The overmigration of (d) can occur either by the use of too high a velocity or if the fault is not perpendicular to the line. The compact mustache of Figure 4 (e), superimposed on genuine reflections in an impossible manner, indicates a small source off the line (in marine work, a shipwreck or extrusion
on the sea floor); migration turns the moustache into horns.
Figure 4
A more extended pattern, such as (f), indicates a fault edge locally subparallel to the line, but again not in the plane of section ( Figure 5 ); the event is a sideways diffraction, having a travel time that must not be interpreted as vertical.
Figure 5
If the fault edge is not parallel to the line, very odd events appear on the section (g). The application of migration to these events is clearly wrong; it finds apparent dips that originate merely in the varying distance of the fault edge from the line, and steepens them, in a manner that is quite false. Figure 3 (d) reminds us that when a fault (or other feature) crosses the line at an angle other than 90 degrees, the correct migration to a sharp edge can be obtained by using a lower migration velocity. If the angle between line and fault is 0, we need to decrease the migration velocity by the factor sin 0. In particular cases, where the details of one fault edge are critical, or where the whole section has the same strike, it may be legitimate to use this artifice. We must remember, however, that the migration velocities then fail to tie at the line intersections. And, in general, the failure of reflection times to tie at the line intersections is made worse.
Fault-Plane Reflection Coefficient Fault interpretation is much more positive if we have fault-plane reflections as well as stratal reflections. Fault-plane reflections are extremely desirable. There are two basic problems. The first is concerned with the reflection coefficient of the fault plane, and the second with the seismic problems of recording very steep dips. Do They Reflect?
Figure 1 [(a) A potential fault trap, and (b) the reflection coefficient of the fault plane] illustrates the problem of reflection coefficient.
Figure 1
In (a) we see a potential trap in a limestone reservoir, and in (b) we see the reflection coefficients produced by the fault offset . . . zero, strong positive, zero, weak positive, negative, strong negative and zero. The length of each reflecting segment, in the figure, depends on the thickness of the layer and the throw of the fault. Significant lengths of zero reflection occur if the layer thicknesses exceed the throw. So, in general, a fault-plane reflection contains positive segments, negative segments, and zero segments. This does not look promising. When we remember that only reflectors of a large area can return strong reflections, it looks even worse. In just two situations might we expect significant fault-plane reflections. The first occurs where the layers are thick and the throws are large. Figure 2 (The type of fault-plane reflection generated when the layers are thick and the throw is large) gives two unmigrated illustrations.
Figure 2
In this situation, the sign of the reflection coefficient may be either positive or negative, and is different for normal and reverse faults. The second does not demand thick layers, but occurs where the throw is so large that the fault juxtaposes layers of markedly different age or consolidation ( Figure 3 , The type of fault-plane reflection generated when the throw is sufficient to juxtapose rocks of vastly different age and compaction).
Figure 3
In this case, the reflection coefficient may be uniformly positive; all layers in the downthrown block are so much younger than their juxtaposed counterparts in the upthrown block that the effect of consolidation outweighs the variations in the layering itself. In fact, fault-plane reflections seem to occur more often than we would expect from either situation. In Figure 4 (Fault-plane reflections generated in a sequence of thin layers faulted with small throw; squinting along these reflections reveal surprisingly constant polarity),
Figure 4
for example, some of the fault-plane reflections seem to be more continuous and stronger, and of more uniform polarity, than we would expect from the sand-shale sequence involved. This raises the possibility that the fault-plane reflection may be affected by more than the simple acoustic contrast between materials across the fault. Specifically, it may be telling us something about the nature of the fault about the width of the fault zone, or the degree of pulverization, or recrystallization under pressure, or cementation, or perhaps even the saturant in the fault zone. This raises our interest, because it contains the possibility of a seismic distinction between a conduit and a seal. Conduits and Seals The fact that faults sometimes act as conduits, over periods long enough to generate large accumulations of hydrocarbons, tells us that the enormous pressures acting to cause intimate closure of a fault are not always sufficient to do so. Therefore, we can visualize the situation shown in close-up in Figure 5 (A close-up of a fault-plane area across which there is intimate contact over only a fraction of the fault-plane),
Figure 5
where the particles that are in contact across the fault, though individually under great stress from the weight of the overburden, make up no more than some fraction of the fault-face area. What is the effective reflection coefficient of this? Everything depends on the saturant. If it is water or oil, the downgoing seismic wave scarcely sees the break . . . the fault-plane reflection is negligible. But if the void is full of gas, the local reflection coefficient of the rock/gas interface is almost -1, while that of the rock/rock contacts is zero or small; the overall effect is a significant negative reflection coefficient whose magnitude depends on the fraction of the faultface area having solid rock-to-rock contact. There is no difficulty in visualizing values for this fraction which would account for a significant reflection. This gas-induced negative reflection does not depend on the throw of the fault or the layer thickness (as do the classical fault-plane reflections we considered earlier). It can persist over the entire length of the conduit with unchanged polarity (though we can expect variations of strength as different rock materials close differently under pressure). Where the reflection is strong, the conduit is very open; where the reflection ceases, the conduit becomes a seal, or the saturant changes to liquid.
Looking again at Figure 4, we have to ask: What other explanation could there be for those fault-plane reflections of apparently unchanging polarity-surely they are gas conduits? And dare we hope that the left-hand fault conduit becomes a seal in its upper reaches, where the faultplane reflection disappears? As we build up a coherent picture in a known area, this type of thinking is no longer fanciful. Such fault-plane reflections do not require any thickness to the fault zone-a millimeter of void will do it. The other possibility for a fault-plane reflection arises if the fault zone is thick-perhaps tens of meters thick ( Figure 6 ).
Figure 6
Then we can visualize a negative-positive or positive-negative reflection doublet from the upper and lower limits of this broken zone, having a time spacing sufficient to retain some amplitude as the reflection pulses interfere. The doublet is negativepositive if the saturant is liquid, but it could be positive-negative if the fault zone is filled with a cement harder than the faulted sediments. However, the reflection coefficients are likely to be small in both cases. A related situation can occur where waters in the fault conduit cause either cementation or leaching in the faulted strata, out to some distance from the fault
(wall-rock alteration). Again the reflection coefficient of this zone is likely to be fairly small; it may also be transitional. Summarizing this discussion, we see three situations in which we expect significant fault-plane reflections: 1. Thick layers with large throws; the fault-plane reflection is likely to be discontinuous and to alternate in polarity. 2. Any layers with very large throws; the fault-plane reflection can be fairly continuous, in which case it is positive for normal faults and negative for reverse faults. 3. Any layers and any throws, but with gas in open pockets in the fault plane; the fault-plane reflection is continuous over the extent of the gas, and always negative.
Steeply Dipping Fault Planes Fault-plane reflections are of value for accurate localization of faults. But fault planes often dip very steeply, and may even approach the vertical in the shallow section (where we are particularly concerned to see the transition from conduit to seal). So we must address the problem of recording steep dips. The following are important considerations: · As always, the location of the lines is the most important variable. · Dip lines must be extended far enough away from the fault to record the fault-plane reflection; this may be a long way. · Strike lines intended to tie the fault-plane reflection must likewise be the appropriate distance from the fault. · To accommodate vertical faults, the group interval should not exceed V/2fm, where V is the velocity to the zone where the fault is vertical (often shallow), and m is the maximum frequency to be maintained through the field work and processing. · In those areas where it is known that the fault dips do not exceed some value q, the group interval may be lengthened to V/2fm sin q. · The group length should equal the group interval; double-length groups are not recommended (either in the field or by mixing or summing in processing). · The stack-array approach should be used. · The simple practical test of all the above requirements is that diffractions having their apices at the fault should be preserved all the way out until their
tails become nearly straight ( Figure 1 and Figure 2 ,
Figure 1
The field work and processing in (a) will produce a good fault-plane reflection; those in (b) will not).
Figure 2
· As always, good resolution is desirable, and this means a high value of fm But any hope of distinguishing the polarity of a fault-plane reflection also requires good preservation of the low frequencies. · However, the terminations of stratal reflections against a fault are best shown by the high frequencies alone (effectively reducing the diameter of the reflection zone). Therefore it may make sense to make one section filtered to pass only the high frequencies, and on this section to provide a red-blue rectified display to accentuate the terminations of the stratal reflections; then we may make a second section, including the low frequencies, to optimize the fault-plane reflection and the reflection correlations across the fault. · The steep dips of fault-plane reflections demand appropriate processing routines; at the least, they demand prestack partial migration (DM0) to provide adequate stacking ( Figure 3 and Figure 4 ,
Figure 3
A fault-plane reflection imaged without and with DMO), and a large-dip poststack migration program.
Figure 4
· These considerations become extremely important in the recurring practical situation where a trap is cut by a fault, the fault provides a conduit from a deep source of gas, but the fault extends to the surface - steepening as it rises. Everything depends on whether the conduit becomes a seal above the trap. If we find a good negative fault-plane reflection at depth, we are encouraged to expect the gas conduit; if this reflection ceases above the trap we are encouraged to expect the seal. But what if the reflection ceases merely because our field work or processing could not handle the steepening dip? · We must seize on those few opportunities that afford an internal check on seismic data. One of these is that, after migration, the fault-plane reflection must fall right on the terminations of the stratal reflections. This is a powerful check-on aliasing, on programs and on velocities. · Because of the need for small group intervals and detailed processing, faultplane reflections are expensive. As always, we have to balance cost and benefit.
Thrust-Plane Reflections At this stage we should consider a few particularities of thrust planes. Obviously, their smaller dip makes them easier to record, without the intensive techniques appropriate to steep dips; thus, the compressional thrust plane of Figure 5 (like its extensional counterpart the glide plane of Figure 6 )
Figure 5
produces visible fault-plane reflections with what we would today call reconnaissance techniques.
Figure 6
In Figure 6 the glide plane maintains constant polarity over a large extent; however, the gliding itself is probably telling us that the underlying material is an overpressured shale, so that there is no message about throw (or conduits or seals) in this maintained polarity. In Figure 5 the zone of constant polarity may be telling us that a conduit exists, and this might be relevant to the prospectivity of the rollover structure. Where there is rollover, of course, we do not need the thrust plane as a seal. But in situations of less intense compression we may see a picture like Figure 7 (a), where the rollover has not yet developed and the fault shows merely as an abrupt change of dip.
Figure 7
In this case we are happy to see the evidence of compression, for while extensional faults rely mostly on chemical precipitation to seal them (which may or may not be there), compressional faults always have the additional sealing mechanism of recrystallization under pressure. Then we have the chance of an additional direct indicator of gas ( Figure 7 (b)); not only do we look for a bright negative reflection over the gas, and a modest positive reflection from the gas/liquid contact, but we hope also for a modest positive reflection from the fault plane within the gas. A Hoax In fault country, many sections (particularly from land work) show low-angle events of constant polarity extending over significant distances, and fortuitously these may look very much like thrust-plane reflections ( Figure 8 (a)).
Figure 8
They are near-surface events reflected from faults whose expression at or near the surface makes a small angle to the line; they may be refractions in a normal nearsurface, or direct reflections if the surface material is thick basalt. The test is to map the events on this assumption ( Figure 8 (c)), and to see if they are coherent from line to line on the map. Of course, each event must be mapped on both sides of its line, since the seismic evidence does not resolve this ambiguity unless either the line or the fault is not straight.
Refracted Fault-Plane Reflections Figure 1, Figure 2
Figure 2
, and Figure 3
Figure 3
illustrate a situation where, for a coincident source and geophone, the downgoing
signal is refracted into a deep high-velocity layer, reflected back at the vertical fault, and "refracted" back to the geophone.
Figure 1
From the geometry and the polarities of the interfaces, the normal-incidence reflection must be positive, and the phenomenon occurs only on the upthrown side; the refracted fault-plane reflection is negative, merging smoothly and tangentially into the negative branch of the diffraction at half the critical distance from the fault. In practice, of course, we have to accommodate CMP geometry, which changes (a) to (b) in Figure 1 (The generation of a refracted fault-plane reflection on a near-trace section). Instead of a triangle of reflection, we have a parallelogram of refraction, but the path lengths above the reflector/refractor are substantially unchanged. Therefore the refracted fault-plane reflection stacks well in its higher reaches, but at later times (where it receives the NMO appropriate to later reflections) the stacking may become poor. The event marked RR in Figure 4 has been interpreted as such a refracted fault-plane reflection.
Figure 4
Figure 3 is the equivalent of (a) for a basin-edge situation where the reflector terminates against a vertical wall of very hard rock. The backward branch of the diffraction is now positive, as is the refracted fault-plane reflection; the two are again tangential at half the critical distance. Figure 5 ( Backward diffraction, fault-plane reflection and refracted fault-plane reflection, from a basin-edge situation) has been interpreted as an example.
Figure 5
A refracted fault-plane reflection is of some value in interpretation. It establishes that there is a high-velocity layer of thickness sufficient to support a significant refraction arrival, and gives its velocity. It establishes positive polarity for the reflection from the top of this layer. It establishes that the layer is faulted by a fault of large throw, probably near-perpendicular to the layers. If the refraction is of opposite polarity to the reflection, it establishes that the fault is down-thrown on the side opposite to the event. And it "points" toward the fault edge (or, strictly, towards the diffraction whose apex is at this edge). Such refracted reflections may be optimized by a separate stack maintaining the NMO (not the stacking) velocity appropriate to the top of the refracting layer; dip may be taken into account. A genuine refracted fault-plane reflection should follow the structure of the refractor. Obviously, it is misplaced (and its refraction velocity falsified) by migration. Occasionally, fault-plane reflections and refracted fault-plane reflections are observed on the same section, as in Figure 5. Figure 6 (a) (unmigrated) is also interpreted as an example; at first glance, we might have expected that the upper
(and stronger) near-linear event would be the fault-plane reflection.
Figure 6
Figure 6 (b) (migrated) shows that it is the lower near-linear event that is the faultplane reflection; the migration of this event is very good. The upper event, though distorted and misplaced by the migration, is interpreted as a refraction along the strong stratal reflection.
Reactivation and Wrenching On rare occasions, the earth surprises us by allowing considerable movement without the reactivation of older faulting ( Figure 1 ) In general, however, the earth cracks along preexisting zones of weakness, so that the reactivation of old faults is very common.
Figure 1
Sometimes we see an example that is clear-particularly when a normal fault is reactivated as a reverse fault. Figure 2 is one such; Figure 3 is its interpretation.
Figure 2
Faults A and B, normal faults after time PP', were reduced in throw, and later sediments reverse-faulted, after time NN'. The direction of the fault plane appears to have maintained itself much the same in the upper sediments.
Figure 3
When the reactivation is in the same direction, the history may not be as clear. Thus a growth fault, for which the activation is continuous rather than episodic, generally yields a smooth plane (which may steepen upwards), and major problems for us in correlating the reflections across the fault. Perhaps the ability to make definite correlations across a fault is our best indicator that the activation was episodic. In this case we can search for the differences in reflection intervals across the fault that would define the times of reactivation, and the amount of throw at each of these times. In this we may be helped by the genesis of new faults at the time of reactivation. Thus in Figure 4 we visualize that the reactivation of the old normal fault engendered a massive slump in the unconsolidated sediments above, along one or perhaps two glide planes angled quite differently from the old fault.
Figure 4
Or, we may expect that antithetic faults are generated at the top of an old fault when it is reactivated, and that when this happens the "continuation" of the old fault is steepened ( Figure 5 ).
Figure 5
In practice, of course, it may be difficult to distinguish this from the commonplace generation of antithetic faults into a curved normal fault ( Figure 6 ); we sometimes get help (as we did in a previous example)
Figure 6
by making the mental distinction between a fault propagating upwards ( Figure 5 ) and one propagating downwards (as in the antithetic fault of Figure 6 ). The reactivation of a vertical normal fault may produce a "fan" of new faults, with both normal and reverse members ( Figure 7 ).
Figure 7
When we see this we are unlikely to understand the details, but the "V" of the fan is still very useful in pinpointing the timing of the episodes of faulting, stability and reactivation. And we always have the hope that what is unclear on one line will become clear as we build the picture from all the lines. We have to accept that the forces involved in successive episodes of faulting (whether compressional or extensional) may not act in the same direction. As we saw in earlier sections, the earth may break into lozenge-shaped blocks, each of which may be rotated-yielding local compressions and extensions, local uplifts and local strike-slip movements. At the grand scale, we are confronted by a wrench zone ( Figure 8 ).
Figure 8
Quite apart from the inherent complexity of two or more systems of faults, we must admit that the seismic method is not very good at detecting strike-slip movement. This sometimes surprises surface geologists, who are accustomed to reading the simple message of offset colors and offset faults on their surface maps. Seismically, at least in the case of layers of constant thickness, there is no direct indication that an observed fault has a strike-slip component. However there are some guides: · First, we positively expect strike-slip faulting if the regional tectonic history is known to include episodes of compression/extension acting in different directions. · We generally exclude major strike-slip displacement on faults that do not involve the basement. If we can see the basement on the seismic section, and the fault extending down to it, we are braced for strike-slip movement; if the seismic quality (or the record length) does not allow us to see so deep, we check the gravity and magnetic data for indication of a basement fault. · We generally exclude major strike-slip displacement on faults that are not vertical or near-vertical. Strike-slip displacement can occur at a small scale on
reactivated inclined faults, but we are unlikely to be able to detect this seismically; it is the vertical faults that put us on guard. · On land, the surface-geological maps may give clear proof of strike-slip offsets at the surface; we must honor these at depth. At sea, just off the coast, we may be able to project strike-slip indicators in the onshore surface geology or the coastline, and may be helped in this by the gravity and magnetic data. · A marked difference of the character of the section across a near-vertical fault is a sure indicator of strike-slip movement ( Figure 9 , A section from an area of wrench tectonics.
Figure 9
Before the wrench, and interesting rollover had formed into a thrust plane (which may well be a gas conduit). This thrust plane (like the stratal reflections) has been offset by the near-vertical wrench fault. A second thrust plane has formed, probably after the wrench). In this we are not fazed by the difference of character that can occur across a growth fault, because growth faults ordinarily dip at a less steep angle.
· Strike-slip movement is suggested wherever zones of intense deformation occur in isolation between undeformed areas. A zone of intense deformation may be seen as a wide vertical swath of no reflections, or chaotic reflections ( Figure 10 , A section across a major wrench fault, illustrating both the zone of chaos (at depth) and the ansence of correlation across the fault).
Figure 10
Because of the broken nature of the rocks in this zone, the forward branches of diffractions from the truncated reflectors are not likely to be seen; on migrated sections, false "smiles" may be generated. A particularly wide swath suggests a zone in which the fault is changing direction in plan, or in which two separated parallel faults overlap. Such wide swaths can be interpreted only by approaching them along clearer faults. · We remember that bends in strike-slip faults (in plan) require the creation of a third direction of faulting; this may show as a distinct fault (typically bisecting the obtuse bend angle), or merely as the wide swath of chaos discussed above. · Whenever older structure is cut by a strike-slip fault, there must be alternations in the sense of the fault, or corresponding variations in the throw; thus, a fault may appear to be here down-to-the-north, there downto-the-south, then down-to-the-north again.
· In practice, it may be difficult to read this simple indicator, because of complication by structure induced just before and during the strike-slip faulting. · The above considerations make it easier for us to contemplate the interpretation of flower structures like those of Figure 11 and Figure 12 .
Figure 11
In Figure 11 (A positive flower structure formed by transpression),
Figure 12
for example, we recognize the steep basement-related fault at depth (which we might pick more nearly vertical than in the illustration), the evidence for sustained extensional subsidence in the thickness of sedimentary section, and the suggestion of strike-slip motion in the incompatibility of intervals between one side and the other; these conjure up a picture of a triangular sliver of rock caught in a compressional pinch between the two sliding rock-masses. We marvel at the ability of the rocks to bend; we conclude that most are shales, and possibly cap-rocks. Then we ask: Are there sands present also, with that fault conduit from below . . . ? But does the structure have sufficient area, in such a depth of water? The stuff of exploration. · In Figure 12 (A negative flower structure formed by transtension), we recognize similar components, but in a situation where the strike-slip movement produces local tension, rather than compression; the sliver of rock drops down into an opening chasm. This is of no interest as a target, but important in telling us that the general regime of Figure 11 -while it certainly has a wrench component-does not necessarily have a compressional component.
· Wherever the nature of a fault allows the interpretation of strike-slip displacement, we search for structural (or even stratigraphic) markers which might show the amount of displacement. Thus the offset of a high ridge (or a graben) across the fault may give us this displacement; so may a line of truncation at an unconformity, or even a well-marked channel. · The other approach, or course, is to identify a reflection interval that changes across the fault, and then to search for the lateral displacement that yields equal intervals. The most positive way of doing this is to construct an interval map. This has the important advantage of highlighting any ambiguity in the reading of the displacement. · It is also worth scanning the structure maps for situations where a longitudinally extended basin is deeper at one end than at the other; faults subparallel to the basin axis may have been subject to strike-slip displacement.
Basic Interpretation Rules From our observations of faults and migration, we can begin to see some basic rules for fault interpretation: · At an early stage, we must form some ideas on the direction of the structural grain and the fault grain, and on how these relate to the line directions. In particular, we wish to identify those lines that are approximately perpendicular to particular faults. Among these, we wish to concentrate the initial interpretation on those that are also on structural dip; these are most likely to be clear. In the early stages we avoid the strike lines, which are likely to be very obscure. · To obtain the fault directions we must correlate the faults from line to line. Often it helps to display the sections at the workstation, in the form of a fence diagram ( Figure 1 ). By rotating the diagram under software control, we can quickly establish the most-likely correlations, and check that no fault splays or other complications appear on the strike lines.
Figure 1
· In correlating faults from line to line, we check that the faults are of the same type, broadly explicable by the same stress mechanism, of the same age, and of compatible throw. In this we are very conscious of the constraints on fault behavior discussed in the previous section; in particular, the displacement ellipse is the control on what we mean by "compatible" throw. · Where fault correlation between two dip lines is ambiguous, we may need to check other lines having less favorable orientations. Thus, the situation of Figure 2 (a),
Figure 2
(The fault picks on the two lines, although made at poor orientation, may permit confident correlation-provided that the sense of the faults is certain: the fault correlation is unresolved), although not ideal for showing the faults clearly, may allow us to work back from the line intersection in the two directions, and be quite confident that the two faults are the same fault. It is important, however, that the faults should be sufficiently clear to leave no doubt about their sense; Figure 2 (b) leaves the two fault directions quite unresolved. · On unmigrated sections, a true fault-plane pick (that is, one made from apex to apex of the diffraction hyperbolas) must tie at a line intersection. This is an important check. · As the picture becomes clear on the dip lines, the fault-plane pick can be carried to all lines, including the obscure lines subparallel to the fault. Again these picks must tie at all the line intersections on the unmigrated sections. So, to a reasonable degree, should the projections of the faulted reflections. ( Figure 3 ).
Figure 3
As shown, on unmigrated sections, and in the absence of gross lateral changes of velocity, all stratal reflections and all fault-plane reflections should tie at the line intersections. Where reflections are truncated before the line intersection, their reasonable projections-within the same fault block-should also tie. Further, correct fault picks should tie at the line intersections. · In general, we use both migrated and unmigrated sections. We use the migrated sections to clarify and locate the faults where the line is nearperpendicular, and we use the unmigrated sections to check the ties at the line intersections until we are confident of the migrations. Where we are fully confident of a fault pick on the migrated sections, we transfer that pick to the unmigrated sections by demigration, with the help of the diffraction apices and use this pick at the line intersections to trace the fault on to the strike lines. The knowledge of the location of fault edges on the migrated dip lines also tells us where to expect sideswipe diffractions (and possibly fault-plane reflections) on the unmigrated strike lines. This is a valuable input to the interpretation of what may otherwise be strange effects on these lines. · As this work proceeds, we decide whether it is better to make an unmigrated structure map, and migrate this map, or to make a map direct from the migrated sections. If we decide to map the unmigrated sections, we know we run the risk of making a wrong reflection pick across a zone
disturbed by diffractions, and of mispicking the faults. If we decide to directly map the migrated sections, we know we shall have a mistie problem, and a need to move updip the location of all lines having a strike component. Each survey has its own appropriate balance between migrated and unmigrated sections; we never use only one, without reference to the other. · We keep in mind the importance of "anchoring" the map at places where there is no dip. · Traps defined by one or more updip faults (either normal or reverse) normally require that we contour the fault planes and the top-reservoir surface independently, in order to calculate the trapped volume. Then we can construct an isometric view ( Figure 4 ), which is particularly useful in checking the reasonableness of the fault/reservoir boundary.
Figure 4
· Faults are often not straight. Likewise, seismic lines on land are often not straight. We therefore face the possibility of seeing the same fault several times on one line ( Figure 5 ).
Figure 5
This may be difficult to recognize, because it happens where the fault indications are blurred by the poor angle between fault and line; nevertheless, it is very important that we do recognize it. We should note also that the apparent age of the faulting may be quite spurious. · The age of the faulting (generally interpreted from the stratal level of the topmost visible displacement) is important to us as a constraint on fault correlation from line to line. However, we have to allow some tolerance on this constraint. In Figure 6 we see a fault plane interpreted from three seismic lines, and its expression (hatched) where it offsets the same reflection (A, at equivalent stratal level).
Figure 6
If now we visualize a contour of the displacement ellipse (such as that corresponding to the smallest visible displacement, as illustrated), we can see that the age pick on the three lines (p, q, r) does not correspond to a constant stratal level. · The behavior of a fault at depth is also useful for correlation, since it contains information on the type of faulting. In one type, the faults (which may be planar) persist to great depth, offsetting the basement ( Figure 7 (a) ); these faults extend downward until the rocks become plastic.
Figure 7
In another type, the fault planes are curved concave-upward, and the faults sole out to bedding planes (or perhaps unconformities) at depth ( Figure 7 (b) ); subsequent adjustment faults may die out at, or just above, these planes ( Figure 7 (c) ). In another type, faults in the shallower section may be absorbed in salt (or other plastic or mobile material), without continuity into the deep section; in the case of salt, these faults may be caused entirely by movement of the salt, but this movement may itself have been triggered by deeper faulting. In the latter case, although individual faults do not persist through the salt, it remains true that a zone of faulting above the salt broadly follows a zone of faulting below the salt ( Figure 7 (d) ). · Significant strike-slip motion usually involves the basement, and is therefore not expected unless the fault is like Figure 7 (a). Further, strike-slip faults are usually near-vertical. These are useful distinctions. · A recurring problem, in the picking of faults, concerns the degree to which we can expect fault faces to be irregular; that is, neither planes nor smooth curves. Often, the joining of reflection terminations on seismic sections seems to need very irregular fault faces ( Figure 8 ), and we are unsure whether to accept these or to insert additional faults. The following observations assist the decision:
1.
Figure 8
Growth faults are generally smooth. (This is Earth's compensation to us for the agonies of trying to correlate across such faults.) We remember that, under conditions of sustained rapid deposition in an extensional basin, the fault faces are curved concave-basinward in plan and concave-upwards in section. 2. On a seismic time section, the usual increase of velocity with depth makes inclined planar faults appear curved, and listric faults appear more curved. Indeed, an inclined fault that appears planar must be slightly convex-up. These effects are not removed by time migration. So, for each survey we interpret, we should take a moment to compute the degree of apparent curvature introduced into planar faults, over a range of fault dips. 3. The division into basement-related and other faults, discussed above, is fundamental. An inclined postdepositional basement-related fault propagates upwards, finding the easiest path; initially the factors deciding this easiest path are rupture strength and overburden density, and these may lead to abrupt changes of fault angle within the section. As movement occurs, these changes of direction may be accommodated by plastic deformation or by pulverization, according to
the type of rock; if not, new faults are to be expected, branching off upwards from the change of direction. On the other hand, faults involved in crestal collapse over an uplift propagate downward, approximately along a radius of curvature; there is little or no accommodation problem, and the faults are likely to be planar. 4. Thrusts tend to follow the bedding in layers that allow sliding (for example, in some shales), and then to step upward quite abruptly in the brittle layers. Seismically, only the "risers" of this stair-step picture may be visible. Identification of the "treads" as shales is feasible, if the migration is trustworthy. 5. Where a fault has become stable, and then sediments have been deposited over it, reactivation of the fault may involve a change of fault direction at the top of the original fault. There may also be an antithetic fault branching off at this point. However, we must be aware of the hoax situation that may arise at an unconformity ( Figure 9 ).
Figure 9
6. The appearance of abrupt changes of fault direction may be caused, erroneously, when faults are themselves faulted ( Figure 10 ).
Figure 10
7. In view of these and other complications, we try where possible to pick fault faces as smoothly planar or smoothly curved, but we are ready to introduce abrupt changes in fault direction where the seismic data quality leaves no doubt, or where we see clear evidence for fault reactivation, or where thrust planes traverse shales, or where we recognize successive episodes of tectonic stress in different directions, or where faults pass close to abrupt changes of rock type.
As the tectonic history emerges, and we recognize the direction and degree of the tectonic forces acting in particular episodes, we can use analog models to guide our expectations and our picking. And, at all times, we remain very conscious of the constraints on fault behavior-particularly the need to conserve volume.
Clearly, fault interpretation can be very labor-intensive, and we cannot afford to study every fault in detail. But we do need to take care with those that represent potential traps. In extensional areas, unfortunately, such traps are often in the "shadow" of a fault ( Figure 11 ), and we must be aware of the problems that this poses:
1.
Figure 11
The fault zone itself is likely to be full of scatterers, decreasing the transmitted signal and adding noise and imperfectly collapsed diffractions. 2. The different velocities along normal-incidence paths a and b distort the reflection times. 3. As shown by the CMP paths c, d and e, the stacking is likely to be impaired by local changes in effective velocity. 4. We particularly need a strike line along the top of the trap, but this is likely to mistie to the dip lines. 5. Depth conversion imposes discontinuities on reflections in the shadow of the fault, leading to doubts about the fluid-content reflection.
6. Strange paths, such as f and g, become possible. If this list appears daunting, we can take some cheer from the fact that we are usually luckier than we deserve. Sometimes we are amazed at the virtual absence of shadow effects, as in Figure 12 .
Figure 12
Often we see only minor effects. Only occasionally do we see major disruptions in the shadow zone, as in Figure 13 .
Figure 13
Incidentally, this figure is of interest as a case where the picking of the main fault is facilitated by noting the numerous gas accumulations (evident both as bright spots and fluid-contact reflections), and using them to infer where the fault must be. The reason for the pronounced fault shadow is almost certainly the heterogeneity introduced by the gas.
Fault Plane Maps A fault plane map is a contour map in which the strike, dip and throw of the fault plane are presented in map view. Remember that fault throw and vertical separation are different; and we must be careful as to which one we are measuring and mapping. Also, keep in mind that fault planes not truly mathematical a plane because they change in strike and dip directions; we sometimes use the term fault surface map to describe the same information as on a fault plane map. A fault plane map is constructed by first measuring the throw (or vertical separation) and depth (or reflection time) of a fault and indicating these values at the proper location on a base map. We generally use either well log and/or seismic data as the source for this information. The common procedure is to post the data values using the notation "throw or vertical separation/depth or time," such as 500/3.200 when the input is seismic reflection time. This example means that the throw is 500 milliseconds and the depth is 3.200 seconds. Of course, we can also indicate (and contour) measurements in feet or meters. However, to map in feet or meters using seismic data, we must have good knowledge of the velocity distribution for an
accurate time-to-depth conversion; we know that the presence of faults in the data introduces complications in the horizontal and vertical velocity distribution. We also are aware that fault shadows create even further velocity complications. Contouring Fault Plane Maps Once the data are posted, they are usually contoured according to depth (or time) only. Sometimes, fault plane maps have both depth and throw contoured on the same map as we shall see later. Contouring fault plane maps requires a different set of guidelines than structural contouring. Some of the more general rules for contouring fault plane maps are (according to Tearpock and Bischke, 1991): · We begin contouring where the data density is greatest. At this point we do not know the fault plane pattern although we can generally tell something from working the data, from regional knowledge, and from glancing at the data values. This starting procedure is contrary to structural mapping where we generally begin at the structural highs and work our way to the structural lows, regardless of date density. · Contours on a fault plane map may be open-ended. Again, this is contrary to structural maps where contours must close (except along the map margins due to the limitations in the size of the map). · Contours are generally smooth except for ramp-and-flat thrust faults and other unusual circumstances where data clearly indicate that changes in strike and dip may be abrupt. · More than one fault plane may be contoured on a single map. Contours, therefore, may split, cross, or merge depending on whether the faults are bifurcating, compensating or intersecting, respectively. Figure 1
Figure 1
is an example of a fault plane map with two sets of contours-depth and vertical separation ( Figure 2 , Diagrams of seismic lines A and B).
Figure 2
In Figure 1 , the fault plane depths are the solid contours while the vertical separations are indicated by the dashed contours. Notice that the vertical separation contours are, for the most part, perpendicular to the fault plane depth contours; this implies that the fault is pot-depositional. This is due to the absence of growth (fault throws are equal for the shallow and deep horizons) on the seismic lines. Bifurcating Faults A fault, generally a normal fault, can split laterally along strike and continue as two smaller faults dipping in the same direction. This is called bifurcation and results in a bifurcating fault pattern. Figure 3 is a block diagram showing the configuration of a bifurcating fault system. We notice in the figure that all the faults are dipping in the same direction.
Figure 3
When the original fault splits into two bifurcating faults, the vertical separation must be conserved. That is, the sum of the vertical separation for the new, smaller faults must equal the total vertical separation of the larger, original fault In our model, shown in Figure 3, we see that the vertical separation of one bifurcating fault (100 feet), added to the vertical separation of the other bifurcating fault (150 feet), equals the vertical separation of the original fault (250 feet). A fault plane map ( Figure 4 ) of a bifurcating system is a plan view of the block diagram shown in Figure 3 .
Figure 4
In the map we see the three points noted in our definition and model: 1. We see that the contours for Fault L intersect the same (value) contours for Fault K. That is, the 12,000 foot L contour intersects the 12,000 foot K contour at the line of bifurcation. 2. The vertical separation of Fault K (±120 feet), added to the vertical separation of Fault L (±65-105 feet), equals the vertical separation of the original Fault L + K (±153-225 feet). 3. All the faults are dipping in the same directionÑin this case, to the southwest.
Compensating Faults Two faults, generally normal faults, dipping in opposite directions can merge along strike and continue as one fault. This is called a compensating fault system. Figure 5 is a block diagram showing the configuration of a compensating fault system. We notice that the two faults, as stated in the definition, are dipping in opposite
directions.
Figure 5
When two compensating faults merge, the vertical separation in the continuing fault must be conserved, just as in a bifurcating fault system. In this case, that means that one of the compensating faults must "take displacement away from," or compensate, the other fault, such that the continuing fault equals the mathematical difference between the two compensating faults. This compensation occurs because of the opposing dip directions of the two faults-faults having the same dip direction add throws along their termination line; faults having opposing dip directions subtract throws along their termination line. In our model ( Figure 5 ), therefore, we see that the vertical separation of one compensating fault (250 feet), less the vertical separation of the other compensating fault (100 feet), equals the vertical separation of the continuing fault (150 feet). Of course, geologically, one fault can not take displacement away from another fault; this is just a simple means to comprehend the mechanics of a compensating fault system. A fault plane map of a compensating fault system based on well data is shown in Figure 6.
Figure 6.
Here we see that Fault B, which dips to the northwest, and Fault A, which dips in the opposite direction (to the south), merge and continue as Fault A (still dipping to the south). Again, we note that: 1. Equal contours of the compensating faults intersect each other along the line of termination (of Fault B at A). 2. The vertical separation of Fault A (100 feet) at depth below the line of termination, equals the difference between the vertical separation of Fault A (300 feet) above the line of termination, less the vertical separation of Fault B (200 feet). Cross section AA' is shown in Figure 7 to aid in the understanding of this example.
Figure 7
Intersecting Faults Unlike the previous two special cases (bifurcating and compensating fault systems), where one fault terminates while the other, in a new form, continues beyond the intersection, intersecting faults both continue after their intersection regardless of their dip direction. This fault pattern may involve both normal or reverse faults. Figure 8 shows a block diagram of an intersecting fault system.
Figure 8
The vertical separation for two intersecting faults may or may not change laterally or vertically beyond the line of intersection. If the two faults occur contemporaneously, then the vertical separation of each fault remains the same before and after their intersection. However, if one of the faults occurs after the other fault (in time), then there will be some compensation of vertical separation across the line of intersection. In either case, conservation of vertical separation must exist. Unless the data strongly support a change in vertical separation across the line of intersection, it is generally assumed that the vertical separation is the same on either side of the intersection. Figure 9 shows a fault plane map for two intersecting faults.
Figure 9
We see that Fault A, noted by solid contours above the line of intersection (to the north) and by dashed contours below it (to the south), dips to the south, while Fault B, noted by solid contours below the line of intersection (to the south) and by dashed contours above it (to the north), dips to the northwest. Again we see that equal contours intersect each other at the line of intersection. The cross section AA' for the intersecting fault system of Figure 9 is shown in Figure 10.
Figure 10.
Here we see that a horst below the intersection, becomes a graben above the intersection. Uses of Fault Maps There are several reasons why we should routinely make fault plane maps as part of the exploration and exploitation processes: · Obviously, fault plane maps provide us with the location of a fault at all depths in plan view. This gives us much more detail than a structural map, which tells us only where the fault is located relative to just one mapped horizon. · Fault plane maps can provide us with strike and dip information at any location along the fault plane. · Fault plane maps give us a better understanding of the three-dimensional structural picture and may even identify new prospects. · Fault plane maps can be very helpful in the design of drilling plans, especially for directional and deviated wells.
· Fault plane maps can be used in conjunction with other structural maps and data to establish the limits of a hydrocarbon reservoir. · Fault plane maps can be used in secondary and enhanced oil recovery projects to postulate the fluid flow pattern.
Structure Maps: Basic Principles Historically, the search for oil and gas has been the search for structure. This is the first message of seismic reflection data. For many years, our task as geophysicists was to find and map structural situations, leaving to the geologist the task of determining what the rocks were likely to be and whether they were likely to contain hydrocarbons. Although our mandate today is much broader, it remains true that our primary function is to delineate structure. Structure is what seismic data does best, with most assurance. Our eye is therefore highly tuned to the recognition of structure on seismic time sections, and to the evidence of whether the structure is genuine. Three questions are fundamental: 1. Is the structure closed in time? in depth? 2. Is the amount of closure in time so great that it would survive any conceivable errors in the static corrections, or any conceivable lateral variation in the velocity used for conversion from time to depth? 3. Is the structure, as evident on the sections and particularly its upper part, displayed in the correct position horizontally? We, as experts, can look at seismic sections and maps, and quickly assess the type of structure apparent from them. We can also look at the annotation of the sections, such as elevations, corrections, velocities and applied processing, and quickly assess the certainties and the doubts about the structure i.e., the measure of the seismic risk, and where that risk lies. In acquiring this skill, it is helpful to have a compendium of many examples. Accordingly, the AAPG has published an atlas of seismic sections in three volumes illustrating the many types of structural style. These atlases, by A. W. Bally (1983), are highly recommended. We enjoy returning to them periodically, and seeing things we missed before. Of course, many of us also keep our own collections of illustrative sections that have come across our desks and have captured our attention for one reason or another. Figure 1 illustrates the Keystone field in West Texas-a classical petroleum trap. As we can see, the field is a simple closed anticlinal dome, with two culminations. The anticlinal axis is shown by a bold line, with arrows to highlight the dips.
Figure 1
We notice that the contours appear natural almost everywhere on this map. The delightful simplicity of the Keystone anticline in Figure 1 occurs only with gentle structures. When the folding is more intense, faulting inevitably occurs. This may take many forms, depending on the totality of the stresses involved. Most anticlinal traps, therefore, tend to look more like Figure 2 -basically a dome, but cut and displaced by many faults.
Figure 2
Indeed, in most exploration mapping it is not possible to separate the operation of contouring from the decisions of fault interpretation. Tying Faults Tying faults means that we correlate the same fault, or faults, from one seismic section to another. We do this for several reasons: 1. It helps us to identify a fault on one seismic section as being the same (or different, if we are incorrect) fault on another section. 2. It helps us to establish, understand and correctly map the fault system within the study area, particularly in distinguishing the major from the minor faults. 3. It helps us to assure that the correlation of horizons and their relative positions across the faults are correct. 4. It helps us to monitor the throw of the faults from one seismic section to another and to assure that we are consistent in our interpretation.
Figure 3 illustrates how we tie faults across several seismic sections.
Figure 3
In the figure, we immediately notice that faults can apparently change dip directions depending on the orientation of the seismic line relative to the strike of the fault. In tying faults, however, we recall that faults on unmigrated sections tie; faults on migrated sections, generally, do not. We use migrated seismic sections to identify and locate a fault, and unmigrated sections to interpret it. Structure Mapping: Example To illustrate how to structurally map faulted seismic data, let us "pick" the following data set. Figure 1 is the shot point map.
Figure 1
Figure 2 ,
Figure 2
Figure 3 ,
Figure 3
Figure 4 ,
Figure 4
Figure 5 ,
Figure 5
Figure 6 ,
Figure 6
Figure 7 , and Figure 8 are the seismic sections, Lines 2-6,11 and 12.
Figure 7
Figure 8
On Figure 9 Line 4, we recognize and mark the fault.
Figure 9
This example is made rather easy for us by three circumstances: 1. We have a train of reflections (3.17-3.30 s at the east end of Figure 9 ) generally stronger than the rest, and with a recognizable character. This makes it easy for us to correlate across the fault, and so to assess the throw of the fault (about 120 ms, down to the east). In the following material we shall assume for the sake of illustration that our potential reservoir is near the top of this train. 2. As we shall see, the fault cuts the line almost perpendicularly. This always makes the seismic reflections more clearly discontinuous and the faults easier to see. In contrast, a line along the strike of a fault can be very difficult to see and interpret. 3. Earlier in this module we reviewed the concept of migration. We remember the importance of this process, which moves dipping reflections an appropriate distance updip, in locating faults correctly. In other modules, we learn that migration also sharpens and clarifies the indication of faults, provided that the faults are approximately perpendicular to the line. In the present case, then, the fault on Figure 9 is particularly clear. This section is a migrated section as are all sections in the figures in this example.
Now that we have recognized the large normal fault as our target, we take our usual broad look at sections. In Figure 10 and Figure 11 , we see a montage of the relevant parts of the east/west sections, aligned approximately on the fault.
Figure 10
As we go north (upwards), the fault develops, increases to a maximum, and then attenuates to nothing again.
Figure 11
We now mark the position of the fault on the lines of the map as shown in Figure 12.
Figure 12.
The temptation is immediately to join the marks together, and to indicate one major fault. However, before things become too fixed in our minds, we must consider the alternative interpretations: is it possible that we are wrong in assuming we have only one fault? So we check on the nearest north/south line (Line 12, Figure 8 ) to see if it has any indication of faults. It has not. We are lucky; we can proceed, at least tentatively, on the basis of the single fault.
Figure 8
So we commence the picking. As always, we prefer to start where the picking is clear. In general, also, we prefer to start on a dip line, because we learn the structural style more quickly, and so can accommodate our picking to it. Where there is faulting, however, we have another consideration: if there is any doubt about the visual correlation of reflections across the fault, and in general there is, we must try to tie our picks around the fault. We pick to the fault on one side, then round on tie lines to the other side, and so confirm or correct the visual correlation across the fault. In the present case ( Figure 12 ), we would like to pick from the west to the fault on Lines 3, 4, and 5, tying our picks to the north/south Line 11, then tie Line 11 to the unfaulted east/west Line 6, and so be able to approach the fault from the east. But unfortunately Line 6 does not tie to Line 11. So we have to use Line 2, to the south. This does not have the big down-to-the-east fault which we are trying to map, but it does have a small down-to-the-west fault. Fortunately, there can be very little doubt about the correlation across this fault. So we decide to start by picking the west end of Line 3, then south on Line 11, east on Line 2, north on Line 12, and so to the east end of Line 3.
Now we have to decide which part of the reflection train to pick. As we look at the lines to the west of the fault (for example, Figure 3, Line 3, near SP 600), we feel reasonably confident that at least two major reflections are present in this train. Let us say that we will pick the upper one. Where is the maximum of its envelope? It is difficult to be sure, but the first major trough between two major black peaks seems a reasonable first guess. Before we proceed, however, we check all the lines to see whether this picking decision will be tenable over the whole area. We conclude that no picking decision is without its problems. In some places we shall need to smooth, or to estimate, or even to ignore small complications which are not relevant to our task of mapping the fault. But we will probably stay with the decision to pick that first major trough. In Figure 13, we pick Line 4 west of the fault.
Figure 13
For present purposes, we ignore the complication between SP 1570 and 1590, being quite confident of our pick on the two sides of it. At the junction with Line 11 we check the tie ( Figure 14 ). This is acceptable, and we turn south. Thereafter we turn east on Line 2, north on Line 12, and west again on
Line 4.
Figure 14
This completes the interpretation of the data. Now we must contour it to produce our structure map. The contouring of the western part of the area is straightforward; mostly, it represents simple regional dip. But the contouring of a fault requires some thought. Generally, one should follow these guidelines when contouring faulted data: · Often it helps to regard the fault merely as a steep dip. Then contours that enter the fault turn and follow the fault, finally emerging on one side or other from the fault. Contours that enter a fault must also leave it. · The throw of the fault follows from the contours on the two sides of it. The variation of this throw along the fault must be geologically plausible. · Contours beyond the end of a fault should show the flex likely to be associated with an incipient fault.
· The spacing of the contours on the two sides of a fault should take into account the likely shape of the surface before it was faulted. Figure 15 shows one version of the contoured structure map. It shows, superimposed on the regional dip to the southwest, a south-plunging anticlinal nose whose east flank is cut by the down-to-the-east fault. It also shows a much smaller down to-the-west fault, which intersects the down-to-the-east fault in a compensating fault system. This assumption may be typical of the judgments which we must make many times a day. There is no escape from making these judgments. However, each of us evolves some personal scheme for marking our working map to show the confidence of these judgments. For example, the continuity of the major down-to-theeast fault across Lines 3, 4, and 5 is highly probable; it is the only interpretation to be seriously considered on the present data. The continuity of the minor down-to-the-west fault across Lines 2 and 3 is possible and plausible, but is certainly not the only interpretation. If it were critical to a prospect, we would need to try several alternatives and consider their risks to the viability of the prospect. It is an important duty of ours to convey to management not only the appeal of the prospect, but also its uncertainty the interpretational risk. We notice the offset between the steep south gradients on the two sides of the fault, and how the implied variations in the throw of the fault are expressed in the contours. If we wished, we could draw a side-view of the fault plane between the two surfaces (for example, a view from the east). The final contour map, of course, is a disappointment to us. We had hoped for a drillable prospect-a closed trap against the fault. Certainly a first look at the sections allowed this hope. But now we find that there is no closure. Both upthrown and downthrown sides of the fault rise steadily to the north or northeast and there is no trap. It should be mentioned at this time, and will be discussed in a later section in more detail, that the development of 3-D seismic data has led to a new perspective on mapping faults. In the past, faults were thought to be long continuous features that gently meandered through the subsurface. This is essentially the same concept as that applied to the previous exercise. With the advent of 3-D seismic data and workstations, this perception is quickly being replaced. The mass of 3-D data has given us the ability to see much more detail than ever before. These new data show that most faults extend only a short distance and that many more faults exist than had been previously thought. Instead of a few long faults, there are numerous short ones. Similarly, 3-D data have changed the way we contour faulted structure maps. As noted above, when using 2D data, contours on both sides of a fault appeared to flow sinuously. 3-D data, however, have shown that contours on the two sides of a fault are not as sinuous or flowing as once believed. Examples of the 3-D seismic expression of faults and associated maps will be shown later. Role of Reflection Time
The message of reflection time includes much more than the location of traps and faults: first, there is the critical importance of reflection time in unraveling the geological history of an area; second, the change in a reflection time interval may also have a message about the type of rock to be expected; and third, the message of reflection time often contains information on the maximum depth of burial of a layer. We will focus on the first of these three relationships. The example we will use is that of Figure 1, from which we will reconstruct the history as shown in Figure 2,
Figure 1
Figure 2
Figure 3 ,
Figure 3
Figure 4 ,
Figure 4
Figure 5 ,
Figure 5
Figure 6 , and Figure 7 .
Figure 6
The key is to draw conclusions from the change of a reflection time interval.
Figure 7
As indicated in Figure 2, Figure 3 , Figure 4 , Figure 5 , Figure 6 , and Figure 7 , there are six major stages of movement and deposition of the seismic section shown in Figure 1. In the first stage (f), the parallelism of reflections suggests stable subsidence and matched deposition from Base Zechstein (Ze) salt through Lower Jurassic (JL). In the second stage (e), truncation of the Lower Jurassic by erosional unconformity suggests uplift, caused by salt thickening in the left half of the section, in Upper Jurassic or Lower Cretaceous time. At this time, some minor faulting occurs below the salt on the left side of the section due to the uplift. Any sediments deposited during these times and before this movement were removed by erosion. Parallelism of the erosion surface and lowest Upper Cretaceous (KU) reflections suggests salt movement had ceased by early Upper Cretaceous time. The third stage (d) is characterized by subsidence and deposition from Upper Cretaceous through Lower Triassic (TRL) time while salt thickening occurred in the right half of the section in the early Upper Cretaceous. During this third stage, minor faulting occurs
below the salt and in the Upper Triassic-Lower Jurassic rocks in the right side of the section due to the uplift. Pronounced salt thickening occurred in two episodes during and after Lower Tertiary (TL) in the fourth stage (c). At this time, the faults on the right side of the section become more numerous. In the fifth stage (b), there is continued uplift, increase in the intensity and growth of faulting and erosion of some of the 'Tertiary sediments. The faulting above the salt now extends through the whole Mesozoic section. Finally, in the last stage (a), salt movement, uplift and faulting ceases. Sedimentation proceeds uneventfully to present day. The conclusions gained from this exercise regarding the structural and faulting context of the prospects are: 1. Minor uplift and faulting begins in the Lower Cretaceous time. 2. Major uplift and faulting occurs in the Tertiary. 3. Faulting of the right-hand structure would probably lead us to prefer to drill the left-hand prospect first. 4. Faulting of the right-hand prospect could block the migration of hydrocarbons to the apex of the structure and trap them downdip where fault closure exists. Contrary to the saying that "timing is everything," in the search for hydrocarbons, timing is only a piece of the puzzle. But it is an important piece that cannot be overlooked. Without knowledge and understanding of the timing of any faulting within a prospect, we run a greater risk of drilling a dry well. Draw a bifurcating fault system in map view where the throw of the original fault is 190 ms before bifurcation and 150 ms afterwards. What is the throw, in ms, of the newly split fault? Draw a compensating fault system in map view where the throw of the two original faults are 200 ms and 70 ms. What is the throw, in ms, of the continuing fault? Draw an intersecting fault system in map view where the throw of the two faults is 150 ms and 75 ms. What are the throws, in ms, of the faults after intersecting, assuming that the faults occur contemporaneously?
Figure 1
Using the fault plane map in Figure 1, if a well is drilled exactly half-way between existing wells number 11 and 14, at what depths would you expect to encounter Faults A and B?
Figure 1
In Figure 1, list the chronological order of faults, from oldest to youngest, with Fault A being on the far left and Fault E on the far right. What type faults are they? Has the area undergone a change in tectonic regime and, if so, what was this change?
Figure 1
The correct answer is 40 ms ( Figure 1 ).
Figure 1
The correct answer is 130 ms ( Figure 1 ).
Figure 1
The fault throws remain unchanged-150 ms and 75 ms ( Figure 1 ). Fault B occurs at -6,860 ft. Fault A occurs at -7,200 ft.
Oldest
Youngest
Fault A
normal
Fault E
normal
Faults C and D
normal
Fault B
thrust
Yes, the basin has undergone a change in tectonic regime. Prior to Fault B, the area was under an extensional stress regime. At the time Fault B occurred, the area changed to a compressional stress regime. Draw a bifurcating fault system in map view where the throw of the original fault is 190 ms before bifurcation and 150 ms afterwards. What is the throw, in ms, of the newly split fault?
Draw a compensating fault system in map view where the throw of the two original faults are 200 ms and 70 ms. What is the throw, in ms, of the continuing fault? Draw an intersecting fault system in map view where the throw of the two faults is 150 ms and 75 ms. What are the throws, in ms, of the faults after intersecting, assuming that the faults occur contemporaneously?
Figure 1
Using the fault plane map in Figure 1, if a well is drilled exactly half-way between existing wells number 11 and 14, at what depths would you expect to encounter Faults A and B?
Figure 1
In Figure 1, list the chronological order of faults, from oldest to youngest, with Fault A being on the far left and Fault E on the far right. What type faults are they? Has the area undergone a change in tectonic regime and, if so, what was this change?
Figure 1
The correct answer is 40 ms ( Figure 1 ).
Figure 1
The correct answer is 130 ms ( Figure 1 ).
Figure 1
The fault throws remain unchanged-150 ms and 75 ms ( Figure 1 ).
Figure 1
The correct answer is 40 ms ( Figure 1 ).
Figure 1
The correct answer is 130 ms ( Figure 1 ).
Figure 1
The fault throws remain unchanged-150 ms and 75 ms ( Figure 1 ).
Figure 1
Using the fault plane map in Figure 1, if a well is drilled exactly half-way between existing wells number 11 and 14, at what depths would you expect to encounter Faults A and B?
Figure 1
In Figure 1, list the chronological order of faults, from oldest to youngest, with Fault A being on the far left and Fault E on the far right. What type faults are they? Has the area undergone a change in tectonic regime and, if so, what was this change?
Figure 1
The correct answer is 40 ms ( Figure 1 ).
Figure 1
The correct answer is 40 ms ( Figure 1 ).
Figure 1
The correct answer is 130 ms ( Figure 1 ).
Figure 1
The fault throws remain unchanged-150 ms and 75 ms ( Figure 1 ).
Figure 1
Using the fault plane map in Figure 1, if a well is drilled exactly half-way between existing wells number 11 and 14, at what depths would you expect to encounter Faults A and B?
Figure 1
In Figure 1, list the chronological order of faults, from oldest to youngest, with Fault A being on the far left and Fault E on the far right. What type faults are they? Has the area undergone a change in tectonic regime and, if so, what was this change?
Figure 1
In Figure 1, list the chronological order of faults, from oldest to youngest, with Fault A being on the far left and Fault E on the far right. What type faults are they? Has the area undergone a change in tectonic regime and, if so, what was this change?
Introduction The concept of structural style is based on comparative tectonics. Its greatest utility lies in identifying certain basic patterns of deformation and faulting that are repeated in geologic provinces throughout the world. Studies of structural style have the considerable advantage of relating these patterns to present-day plate tectonic habitats and thereby to predictive models of origin and evolution. This means that prior to exploration, the types of source and reservoir rocks, migration paths and hydrocarbon traps can be predicted, or at least anticipated. In classical structural geology, surface mapping, well data and laboratory experiments were the main sources of information for understanding tectonic features and their genesis. This has changed dramatically within the past several decades. Seismic reflection profiling has provided us with a tool that can reveal continuous structural data to considerable depths. It has opened up both submarine provinces and the deep continental subsurface to eager scrutiny.
As developed by Harding and Lowell (1979), Bally et al. (1983), Bally and Oldow (1983), and Lowell (1985), structural styles are first differentiated on the basis of whether basement is involved or uninvolved directly in deformation. In the case of noninvolvement, structures primarily develop within a "detached" sedimentary cover. These two basic criteria are, in turn, applied to the four major types of tectonic provinces-compressional, extensional, strike-slip (or wrench) and cratonic (or vertical). For the most part, these last two types of province are thought to involve basement in nearly every case. First of all, we need to define what is meant by basement. For petroleum geology, this is usually taken to be structural basement, which means rigid crystalline igneous or metamorphic rock. The degree and manner of its involvement in the creation of structures are important to understand, since these determine the overall context for structural entrapment in overlying sediments. In fact in many hydrocarbon provinces, basement structure is the key to both overlying deformational and depositional patterns. These provinces and the structural styles within them are related genetically to one or more of the following plate boundary classifications: 1. divergent, which includes intracontinental rifts, proto-oceanic (new ocean) basins and oceanic basins 2. convergent, which includes intra-oceanic arc (oceanic/oceanic) systems, continental margin arc (oceanic/continent) systems, closing ocean (continent/continent) basins and both forearc and backarc zones 3. strike-slip, which includes any of the above basinal systems where the dominant deformational forces are oblique to each other rather than in the same plane and the resulting movement is predominantly horizontal rather than vertical Likewise, just as basins and structural styles are a product of a tectonic process, faulting within a basin is generally a product of that same tectonic process. Thus, the dominant type of faulting within a particular basin is the result of the dominant deformation force. The classification of structural styles-whether referring to the overall basin, the individual structures or the faults-is in fact the same. Divergent Basins: Overview Divergent or rift basins are closely linked with plate divergence and seafloor spreading. These basins generally form as down-faulted failed arms, where continental plates are stretched but do not completely separate. They are usually fairly small, linear basins that can be either deep or shallow ( Figure 1, A typical rift basin. Generalized cross section through the Suez Basin of Egypt). Horst-and-graben tectonics generally give them an irregular profile.
Figure 1
Not all divergence results in failed rifts. Some rifts will evolve into pull-apart basins, as the seafloor continues to spread and basaltic ocean crust is added ( Figure 2, A typical pull-apart basin.
Figure 2
Generalized cross section through the Gabon Basin of west Africa). These basins form next to the trailing edges of continents that have been split apart. For this reason, they are also sometimes known as passive margin basins. Pull-apart basins are generally long, linear, asymmetric basins with a one-sided continental source. They occur mostly offshore on intermediate sub-basin crust. Most of the world's pull-apart basins are located on both sides of the Atlantic and Indian Ocean margins.
Divergent Basins: Basement-Involved Faulting Regional block faulting is perhaps the most widespread structural style in the earth's crust, characterizing a majority of the world's passive continental margins ( Figure 1 ),
Figure 1
as well as numerous, linear intracratonal grabens that occur on nearly every continent ( Figure 2 ).
Figure 2
In most cases, block faulting is directly related to the process of rifting (i.e., continental breakup). In a few possible exceptions, e.g., the Basin and Range province of western North America, crustal extension still demonstrates the principal features of rift tectonics and thus may, in fact, represent unique or anomalous circumstances of this same basic process. It should be noted, however that block faulting also occurs along some convergent plate margins and in backarc settings. Tectonism in these provinces may be thought of as dominantly tensional in nature. At the same time, rifting is a complex process; rarely, in fact, during its initial stages does separation occur exclusively at right angles to the central rift axis. Components of wrenching and compression are common; the former may be related to transform offset of ridge segments. individual faults, or fault assemblages, may therefore show considerable strike-slip displacement. Local folding and reverse faulting may also exist. Perhaps the most important characteristic regarding normal faulting is its deceptive simplicity. Though normal faults most often appear relatively simple in cross section, they are most often extremely complex in map view. Individual faults can be straight or cuspate, or can alternate between the two. Individual blocks bounded by faults
can vary substantially in size and can be tilted or rotated in different directions to different angles within a few square miles. In places, faults may appear to die out in ways that seem contrary to the predictions of geometry or rock mechanics. In a broad sense, much of this can be thought of as resulting from the intricate readjustment that occurs due to the creation of "extra space" by extension. In many cases, the precise pattern of structure related to a particular prospect or play may not become clear until years of detailed work have been done. Figure 3 shows three basic models that have been proposed to explain the geometry of rift-related normal faultingÑstraight, listric and domino. Mach controversy continues to surround this subject.
Figure 3
The degree to which basement-involved faults show listric geometry appears to vary between regions. Such variation is believed to result from differences in how rifting evolved and in the type and thickness of crust involved. Due to the excellent exposure and relative youth of its features, the Red Sea graben is often used to point out the principal aspects of this structural style. In a study of the evolution of the Red Sea ( Figure 4 ), Lowell et al.
Figure 4
(1975) recognized that the larger, regional faults vary in strike and frequently intersect, but demonstrate a dominant orientation that parallels the central graben. Two important secondary fault trends appear to form a conjugate set that is bisected by the primary trend. Displacement along them, however, is almost entirely dip-slip. Both trends appear to have developed simultaneously during various stages of rifting. Together, they represent the principal style of rift-related faulting. Active extension, and thus faulting, appears to have been episodic. As a result, considerable overprinting of older structures by newer ones has occurred. The cumulative effect is great complexity on a local scale, despite the overall consistency of major structural trends. Most faults in the Red Sea graben are interpreted to show listric geometry at depth in cross section ( Figure 4 ) and a curving trace in plan view. Seismic profiles, however, have been interpreted to show more or less straight, relatively high-angle normal faults, at least in several areas. Since reflection profiles generally exaggerate the curved geometry of faults at depth due to velocity increase, the linearity of these fault planes appears all the more distinct. The question of detailed fault geometry, therefore, remains partly unanswered in this region.
In basement-involved divergent basins, we often see fault-bounded blocks rotated in stair-step fashion toward the central graben axis. Seldom, however, is the structure this simple. Antithetic faults, those that dip opposite to the major faults (i.e., away from the rift axis), are also common. These, too, can be thought of as conjugates. Fault wedges and basement blocks can be rotated in any number of directions, and both the geometry and magnitude of displacement often vary considerably, even between adjacent blocks. Sufficient rotation can cause reverse drag folding in hanging-wall sediments, such that beds "roll over" and thus imitate the structure of many growth faults. In addition, a phenomenon known as footwall uplift, where the footwall is physically displaced upwards as a result of isostatic and elastic rebound, has been identified as accounting for about 10 percent of the total offset along basement block faults (Jackson and McKenzie, 1983). Such uplift can lead to local thickening of sediments over the hanging-wall block, thinning over the footwall block, and to local bathymetric highs that can localize reef growth or other shallow marine sedimentation. A look at the Piper Oil Field in the central North Sea graben ( Figure 5 ) reveals several other major features consistently seen in basement-involved divergent basins.
Figure 5
Figure 6 ,
Figure 6
Figure 7 ,
Figure 7
and Figure 8 are taken from Maher (1980) and show the relationship of production to block fault structure in the Piper Field.
Figure 8
Note the location of the oil/water contact. Given that migration and entrapment most likely occurred after faulting and that erosion tilted and sealed the Piper Formation, the faults within the sand body are probably nonsealing. Cross section ( Figure 7 ) and seismic data of the Piper Oil Field ( Figure 8 ) reveal how strata are draped over basement blocks. These blocks exercised a dominant control over depositional environments. In particular sedimentation of the highenergy, marginal-marine Piper Sand was largely determined by the location of structurally positive horsts. Moreover, later movement along faults created migration paths into the sand body. Thus, rifting was instrumental in creating conditions conducive to both the deposition and the favorable structural geometry of source and reservoir rocks. Basement block faulting also occurs at plate convergence and in association with transform faults. This structural style characterizes backarc basins and sometimes the inner margin of forearc basins. To date, the former have been found to be more petroleum productive. With regard to transforms that cut through continental crust, such as the San Andreas, strike-slip motion sometimes generates extension and normal faulting on the cratonal side of related basins. In the southeastern Great Valley of California, such faults were generated during the episode of most rapid movement along the San Andreas, and have acted to trap significant amounts of petroleum.
Passive continental margins represent one side of a successful rift separation. The term passive refers to the general assumption that most orogenic tectonism has ceased along such a margin. They are characterized by seaward-pro-grading wedges of mostly shallow marine sediments that can reach 14 km in thickness (Bally and Oldow, 1983). In their overall character these deposits reflect the basic stages of rifting and continental separation. Basically, then, the evolution of passive margins can be divided into an early, riftrelated phase and a later drift-related phase. Earlier structures are predominantly associated with basement block faulting. Later phase deformation is concentrated in the overlying sedimentary wedge and is controlled mostly by gravity (e.g., growth faulting, salt and shale structures). A more detailed look at these two basic structural levels is offered in Figure 9 and Figure 10 , which shows the current configuration of the passive boundary along northwest Africa.
Figure 9
We note, in this case, how three sedimentary sequences are evident.
Figure 10
In this figure, we see that the early phase lithotectonic sequence is confined to the Triassic-Lower Jurassic units, while the bulk of the later, continental margin phase is represented by Cretaceous and Tertiary sediments and that the structures developed within them. With regard to structure, where the Triassic-Jurassic sequence is characterized by basement faulting, the Cretaceous sequence shows large-scale detachment faulting, and the Tertiary sequence is, so to speak, faultless. We might notice how the detachment faulting appears to be broadly related to deeper basement block faults. These, however have not been reactivated such that they cut up into the Cretaceous sequence. When we compare the seismic section of offshore western Africa shown in Figure 9 and Figure 10 to that of the Baltimore Canyon area (offshore U.S.) shown in Figure 11, we immediately see the similarities in the faulting style.
Figure 11
What we are seeing are mirror images of two sides of a seafloor-spreading center. As in western Africa, the Baltimore Canyon area is dominated by deep, basement block faulting with younger detached gravity-slide deformation. Numerous types of hydrocarbon traps occur in basement-involved fault blocks in rift provinces and passive margins ( Figure 12 ).
Figure 12
Deep burial may, however, modify some of these, as will salt mobilization. The more important role of this tectonic setting, vis-a-vis petroleum potential, has been to act as the site for the deposition of great clastic wedges, especially deltaic, that are characterized by relatively shallow detachment faulting, as well as salt, shale and gravity sliding structures that have trapped enormous amounts of hydrocarbons.
Divergent Basins: Growth Faults By far the greatest number of hydrocarbon traps of the non-basement-involved type have been found along passive margins in association with growth faulting, so named because of its syndepositional nature. They are best developed in regions of deltaic or continuous clastic shelf deposition, where sediment accumulates rapidly and remains semi-consolidated to depths of several kilometers. High pore pressures in shales due to retarded dewatering is a frequently cited influence on the generation of growth fault structures. Regional crustal subsidence due to sediment loading provides at least a broad context for extension. However, a great many growth faults are also directly associated with salt diapirs, which essentially act as large, locally upthrown blocks. Historically, the Gulf of Mexico has served as the type locality for this structural style. Figure 1 shows an offshore growth fault, as interpreted on a seismic profile.
Figure 1
Some of its more important characteristic features are: 1. The major fault plane is listric in shape, dips into the basin, and becomes parallel (flat) to bedding with depth. 2. Displacement along the plane increases with depth. 3. The thickness of individual units increases abruptly in the hanging wall, due to syndepositional offset. 4. Both secondary synthetic and antithetic faults complicate the basic structure. Many of the former sole out along the main detachment plane. 5. Back rotation of the downthrown block creates rollover that increases, then decreases and shifts basinward with depth. Some of the secondary faulting appears to be related to extension in the crest of this structure. 6. In map view, the fault plane is cuspate, and is usually concave toward the basin center. Antithetic faults, however often show the reverse geometry. The flattening of faults with depth has been variously interpreted. Explorationists have traditionally pointed to the existence of overpressured, ductile shale intervals into which the faults often seem to die out. Roux (1977), however has attributed this flattening to increasing compaction-faults originally develop at steep angles that flatten with burial as bed thickness decreases. Of these two interpretations, the former continues to be preferred. Shale units in which the effective in situ stress is especially low can be assumed to be preferential zones for faulting. Pore pressure effects ensure very low shear
resistance. This preference also appears to influence fault geometry. As shown in Figure 2, faults flatten out into shale-dominant prodelta slope facies.
Figure 2
The pattern of such faulting can be directly or indirectly related to preexisting basement structures, as we have mentioned. The rejuvenation of deep-seated faults by sediment loading can directly lead to structural adjustment in the overlying cover. Preexisting basement structure can more indirectly establish the context for growth faulting by influencing the development of depositional hinge areas. This, in turn, can directly affect the location and trend of detachment structures. An example of indirect basement influence is given in Figure 3, which shows a complex example of large-scale, growth-type faults developing above the deepest portion of a rifted margin.
Figure 3
In addition, deep-seated faults can be rejuvenated by sediment loading and this may cause structural adjustment in the overlying cover. Here, older, fault-generated basement topography has determined thickness and facies patterns of later sedimentation and, thus, the location of highly ductile limy shale intervals. Flowage of this lithology under the influence of gravity-induced stress has played a significant part in the development of growth faulting ( Figure 10 ).
Figure 10
Some researchers consider gravity to be the principal cause of growth faulting in areas such as the Gulf of Mexico or Niger deltas. With respect to gravity, the maximum principal stress is vertical. The shear component of this stress, as resolved along the regional depositional slope, leads to instability. In general, body forces within a volume of sediments deposited over a sloping surface must be considered significant. Where this volume is very large, as in the case of passive margin clastic wedges, these forces are very likely to be fully capable of generating large-scale fault systems. At the same time, a variety of other tectonic influences, related to salt and shale diapirism, possible basement fault rejuvenation, and so forth, exist in most passive margin settings. Since growth faulting seems to be a regional response to overburden and slope-related instability or metastability, it may result from a number of specific causes. Figure 4
Figure 4
and Figure 5 of the Hibernia anticline show the fallacy of trying to strictly segregate structural styles in every situation.
Figure 5
The Hibernia discovery was made approximately 250 miles due east of Newfoundland in a narrow, Late Mesozoic graben in the Jeanne d'Arc Basin that cuts into the northern portion of the Grand Banks platform. The graben developed as part of the active rifting phase that separated Africa and North America between the Late Triassic and Cretaceous ages. The structural style, at least in Lower Cretaceous beds ( Figure 5 ), is growth faulting with large-scale rollover. Compare Figure 1 to Figure 4. The faults, in Figure 4 case, developed as part of the rifting episode and though syndepositional to some extent, are not all detachment structures. The more major faults extend deep into the Paleozoic section and most likely continue into basement (Arthur et al., 1983). Hibernia, therefore, presents a case in which the geometry of features assigned to a non-basement-involved structural style of growth faulting also appears in the features of a basement-involved rift style of block faulting. Figure 6, Figure 7
Figure 7
, Figure 8
Figure 8
,
Figure 6
and Figure 9 shows some of the more common hydrocarbon traps that result from growth faulting.
Figure 9
Divergent Basins: Salt Movement Faults Extension above salt structures causes anticlinal warping and normal faulting associated with crestal collapse that usually occurs with shallow penetration. The latter is sometimes termed keystone, for reasons that should be apparent in Figure 1.
Figure 1.
With depth, however faults associated with salt structures are more influenced by the upward movement rather than by downward crestal collapse. As such, normal faulting still dominates, but most often appears radial in its pattern. As with normal faulting in other structural styles, local complexity around the salt features can be great. Antithetic faults are very common. The basic scheme of hydrocarbon traps associated with salt structures is shown in Figure 2.
Figure 2.
In many cases, the largest traps are in reservoirs not pierced by the salt, as in, for example, numbers 1, 2, 8 and 9 in the figure. This, however, is not a hard-and-fast rule.
Divergent Basins: Shale Movement Faults
The conditions that encourage the flowage of shale within a thick wedge of clastic deposits are primarily related to retarded dewatering of shale and to the overall sequence of sediments in a regressive (progradational) shelf. During rapid burial, sands dewater and compact faster than do shales. Impermeability acts to inhibit water loss. An exception to this occurs at the contact between a shale interval and an overlying porous sand, but here the expulsion of pore water into the sand creates an upper seal even more impermeable to continued water loss from the underlying shale. Trapped, excess pore water keeps the clay mineral grains apart, thus countering lithostatic pressure, and keeps the rock in a weak, ductile state, susceptible to flow. This is the classic explanation for the overpressured shales so commonly (and, often, unfortunately) encountered along passive margins buried by thick, regressive clastic sediments. Along with Jackson and Galloway (1984), we might think of the entire depositional body as a "spoon" ( Figure 1 ), composed of a proximal, sand-dominated "megafacies" wedge that overlies a more distal, shale-dominated one.
Figure 1
This setting leads to what is known as differential compaction. As the sands compact and become more dense, they subside into the less compacted shales. Experimental and theoretical work have both shown that the stress resulting from differential compaction and loading is greatest at the front of the load and that the more mobile, overpressured material beneath will flow laterally to allow continued load subsidence. In geologic terms, this means that, due to the depositional slope, shales will be forced basinward. This can lead to growth faulting on a local scale and the creation of shale mounds. Such mounds can subsequently act to localize sand deposition, as progradation continues, and, thus, the generation of other growth faults ( Figure 2 ,
Figure 2
Figure 3 ,
Figure 3
Figure 4 ,
Figure 4
Figure 5 ,
Figure 5
Figure 6 , Figure 7
Figure 7
, and Figure 8 ).
Figure 6
They may also rise as diapirs into overlying units or they may spread laterally.
Figure 8
Differential compaction apparently leads to the transfer of great masses of shale, sometimes ahead of a series of growth faults. Bruce (1973) has described shale ridges in south Texas with dimensions of over 20 km in length, 40 km in width, and 3,000 m in height, with a trend that is roughly parallel to regional strike. Shale diapirs also seem to be related to gravity sliding along master growth faults, which become planes of detachment at depth. Shale structures are not usually of great significance to petroleum exploration. The Beaufort Sea represents an important exception to this observation. Shale structures are, however, intimately associated with growth faulting, salt diapirism, and the overall structural evolution of thick clastic wedges along passive continental margins. This makes knowledge of them necessary to both regional and local exploration in such settings.
Divergent Basins: Gravity-Induced Faulting Gravity-induced downslope movement has been invoked by structural geologists to explain an almost dizzying variety of phenomena on almost all geologic scales. It has been used, for example, as an interpretation of the rumpled-cover nappe structures of the Alps, the de'collement thrusting of the Canadian Cordillera, the chaotic sedimentary sequences known as melange or olistostromes, the disturbed layering of single depositional layers, and the generation of growth faulting. Gravity is now invoked to explain specific small scale structures. In many instances, pore pressure effects are thought to play a primary role in initiating actual detachment. Tilting due to tectonic uplift will create slopes along which the force of gravity will be resolved as shear stress. This has been applied to specific decollement structures (see the classic explanation of the Heart Mountain Fault by Pierce, 1957), and also to both local and regional detachment features along Gulf Coast-type passive margins characterized by high rates of sedimentation. Among the latter are rotational slides observed high on upper delta front slopes and larger-scale, slowermoving rotational blocks that show growth fault-type structures at their head. A third common setting for apparent gravity-related displacement is in thrust belts, where topography can become oversteepened due to the continued advance of thrust sheets. Here, local normal faults will relieve the resulting stresses, which are almost entirely due to the simple weight of the material involved (i.e., the maximum principal stress is gravity and therefore, vertical). In relation to normal deposition along passive, divergent margins, so-called thinskinned gravity sliding is seen to occur in close association with growth faulting and shale diapirism (Jackson and Galloway, 1984). 'Two proposed versions of the structural evolution which might be responsible for this close interrelationship are shown in Figure 1 ,
Figure 1
Figure 2 ,
Figure 2
Figure 3 ,
Figure 3
Figure 4, Figure 5
Figure 5
, and Figure 6 .
Figure 4
These have been derived from detailed study of the Mexican Ridges area in the western Gulf of Mexico.
Figure 6
The difference between the two models lies in the steepness of the shelf slope. Zone 4a, which has a slightly steeper slope, is characterized by gravity sliding, while Zone 4b, which is characterized by growth faulting, has a slightly flatter slope. In the view of most researchers, each glide sheet in this type of environment has a growth fault at its head, a decollement along its principal length, and a thrust or faulted fold at its terminus. To some degree, these sheets can be likened to giant
slumps, with growth "s carps" marking the upper planes of rotation. We should notice the ponding of sediments between and over the ridges. The loading from this accumulation is thought to effectively stop gravity gliding and help initiate shale diapirism. Figure 7 shows an excellent example in seismic section of a productive rollover trap (Slick Ranch Field) associated with such a glide sheet.
Figure 7
The irregularity of the glide plane, which is partly due to shale diapirism, acts to localize the degree of bed rotation. Movement along the basal detachment is assumed to be by continuous gravity creep. In the section we note that the top of the Vicksburg Formation becomes conformable in the direction of gliding. As with shale structures, these gravity-induced features are important to petroleum exploration mainly for their relation to growth faulting. The anticlines caused by the type of downslope movement shown in Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , and Figure 6 . should be considered potential hydrocarbon traps in deeper water areas off passive margins. Melanges, on the other hand, are important to onshore exploration, since they directly indicate the facing direction of a fossil arc, and thus the general orientation of structural trends to be expected.
Convergent Basins: Overview There are four basin types that form when plates converge near continental margins. These are the forearc, backarc, strike-slip (or non-arc) and collision basins. All four convergent margin basin types have a number of features in common and will be discussed to some extent as a group.
There are two types of basins that are found near subduction zones that have developed island-arcs. Backarc basins form between an island-arc and a continent. Forearc basins lie between the island-arc and an ocean trench. Non-arc basins are formed along convergent margins where the plates move by transcurrent faulting. Consequently, they are often called strike-slip basins. Strikeslip basins are small basins that form through a combination of both the transcurrent fault movements and local block faulting. Strike-slip basins include the Los Angles and Ventura Basins of California, the Vienna Basin, and the Crimea and Baku Basins of the Soviet Union. Collision basins, sometimes called median, intermontane or successor basins, are small basins formed within marginal fold-belts, along sutures where either two continents, or continental coastal mountains and a trench, have collided. Some explorationists consider this a type of backarc basin, but a well-defined trench-arc system is usually no longer present.
Convergent Basins: Basement-Involved Faulting The two principal elements of basement-related structural styles are compressional fault blocks and their bounding basement thrusts, which may range in inclination from near-vertical to less than 30 degrees. Both high- and low-angle faults occur along a single thrust front and are now thought to represent the end members of a continuous spectrum of basement-involved deformation. Reverse faults involving basement generally occur along convergent plate margins, primarily in foreland regions, and in the inner trench slope and outer high portions of certain subduction zones. Of these, the foreland regions are by far the most important to petroleum exploration. Basement-involved forelands primarily develop in two major settings between a volcanic arc and a craton (i.e., in a "backarc" setting), and in front of the arc during continental collision. The first of these is particularly characteristic of the North and South American cordillera, while the latter typifies the great Alpine/Himalayan orogenic zone. Compressional basement block faulting is particularly well known and well studied in the Rocky Mountain foreland region of the western United States. In this area, large, rigid masses of Precambrian crystalline rock have been forced up and, to some degree, laterally outward many thousands of feet. The larger uplifted blocks appear to be intimately associated with deep foreland basins. In general, each block has its related basins, whose timing of subsidence seems directly related to uplift history. Most often, these basins are strongly asymmetric, with their structural axes running close to the thrust front and their back flanks forming relatively gentle basement slopes away from the uplift. These basins have served as the sites for considerable sediment petroleum generation and entrapment. Structural relief between the present crest of basement uplifts and their associated basins is frequently on the order of 10 km or more. Thus, the vertical component to faulting is undoubtedly considerable. The overlying sedimentary cover on these blocks has usually acted in a passive manner,
accommodating basement uplift by drape folding and brittle fracture. Blocks are typically bounded by faults on both sides and are tilted at various angles. Figure 1, Figure 2
Figure 2
, Figure 3,
Figure 1
Figure 3
and Figure 4 show four styles of faulting that have been proposed to explain the geometry observed in the field and in seismic profiles.
Figure 4
To some degree, the differences between these reflect the fact that various investigators have concentrated their efforts in different parts of this region, along different major fault zones. Such zones display variable geometries, which range from reverse listric planes (steepening with depth) to low-dipping thrusts of 30 degrees or less ( Figure 6 ).
Figure 6
Single faults apparently show both these geometries along their strike. At present, the proposed styles of faulting reflect two major schools of thought. Both schools agree that compression is involved to some degree in deformation, but part company on the question of whether displacement is dominantly vertical (upthrusting of Figure 1 and Figure 3 ) or horizontal (over-thrusting of Figure 3 and Figure 4 ). The seismic profiles given in Figure 5 and Figure 6 illustrate how apparently irreconcilable these two schools are.
Figure 5
Many of the principal faults have also apparently suffered a degree of ancillary strike-slip displacement. For these reasons, we now believe that at least several of the major blocks, such as the Wind Rivet have been uplifted in a type of scissor-like rotational pattern, with the inclination of the fault plane decreasing and the amount of lateral movement increasing with distance from a "nodal" zone of almost total vertical uplift. This has, in turn, been related to the clockwise rotation of the entire Colorado Plateau during Laramide time (Gries, 1983). With respect to those blocks bounded by near-vertical faults, Harding and Lowell (1979) describe three basic structural levels, from basement up into the sedimentary cover: 1. a tilted fault block of basement and immediately overlying units 2. an intermediate level where the principal fault becomes a zone of steep drag folding 3. an uppermost level of gentle drape of monoclinal folding The structure of fault segments can be complex, involving normal faulting along the crest of the block, drape fold, tear faulting of the cover and basement, and more minor reverse faulting in the footwall of the main fault.
Some of our colleagues have pointed to older zones of Precambrian shearing within the basement itself as a probable control on location and possibly geometry of faults. Certainly, in any region where basement faulting is a dominant structure style, we should understand (or attempt to understand) the preexisting structural character of the rocks involved. Metamorphic rocks are, in general, the products of orogenesis and very often contain important planes of weakness, such as shear zones and major lithologic boundaries, that will have some amount of influence on all later deformation. Another possibility involves the inversion of rift-related block faults during later compressional tectonism. Convergence of older passive plate margins can lead to the rejuvenation of originally normal faults or thrusts. Such reactivation, since it reverses the sense of displacement, is said to create inverted basins. Understanding this can be crucial in certain regions, since the sediments normally assumed to characterize rift provinces will become involved in anticlinal, thrust, and compressional-wrench tectonism. In such cases, apparent horst-and-graben morphology may be retained; however, the more recent uplift will have substantially altered the original structural configuration. Figure 7 indicates some simple hydrocarbon traps associated with basement-involved thrusting.
Figure 7
Convergent Basins: Non-Basement -Involved Faulting The primary structural style in this category is the decollement foreland thrust-fold belt. Due to the thorough work of such authors as Price and Mountjoy (1970), Dahlstrom (1970), and Bally, Gordy and Steward (1966), the Canadian Cordillera has come to be generally treated as a type locality for such deformation. Figure 1 is the classic cross section by Price and Mountjoy (1970) through the southern portion of this structural belt and shows most of the principal features we have come to expect in this style.
Figure 1
The majority of hydrocarbon production exists in the foothills region, where faulting is especially complex. Thrust-faulted anticlines appear to be the most common traps in productive foreland regions throughout the world. But with regard to the size and specific geometry of traps, it should be emphasized that considerable variety exists within and between these regions. Over the span of Proterozoic time, differences in the size and morphology of plates, in their marginal sedimentary character, inherited structural features, and specific motions have all ensured a high degree of structural variation in foreland orogenic zones. Like compressive basement faulting, decollement thrusting is most often related to processes occurring along convergent plate boundaries, both in the mentioned foreland belts and along the leading edges of subduction zones. Figure 2 displays the interpreted structure on a seismic section across the active subduction complex offshore of northern California.
Figure 2
This section shows the de'collement faulting in the accretionary prism quite well. We note the change from the chaotically deformed accretionary prism sediments into the more conformable forearc basin at the far right in the figure. The deformed complex is broken into thrust slices about 400-500 m thick, whose bounding faults display variable curvature. This is probably due to a degree of stratigraphic control and ramping of fault development. We can also see that where compressional features dominate the accretionary prism, extensional faulting characterizes the deep sea sedimentary section west of the trench. This has been observed in a number of active subduction zones. Often, these faults have developed in the underlying ocean crust and have been interpreted as being the result of lithospheric bending and consequent tensional rupture. Collision-related thrust belts ( Figure 3 ) verge toward the subducting ("oncoming") plate.
Figure 3
A transition to basement-involved thrusting often occurs in these settings behind the thrust front. Figure 4 shows how the involvement is usually interpreted to increase into the mobilized, metamorphic "core" of an orogen, such as the Canadian Cordillera.
Figure 4
In the case of trench deformation, the situation is less well-understood. 'Two forms of basement-involved faulting are presumed-that which incorporates "slivers" of ocean crust, known as ophiolites into the accretionary prism, and that which cuts continental crust. For both trench and foreland thrust zones, overlap occurs between detached and "connected" structural styles. Most thrust belts presently exposed at the earth's surface occur as sinuous belts up to thousands of kilometers long. Their width is not uniform, but is instead characterized by sharply recessed and more gently extended portions known, respectively, as reentrants and salients. The cause for this type of variation along tectonic strike is not well-understood, but often thought to be related to preexisting basement features. In discussing foreland thrust belts, there are a number of general aspects that are of direct importance to hydrocarbon exploration: 1. Anticlines occur primarily in the hanging wall, and are asymmetric toward the direction of tectonic transport. 2. The geometry of thrusting is much dependent on the competence of the sedimentary units involved. The number of thrusts, as well as the intensity of folding, decreases in distinct proportion to increases in the amount of thick, competent units (e.g., sandstone, carbonates). 3. Sedimentary thicknesses decrease and deposits change character to more shallow-water clastics in the direction of tectonic transport, i.e., away from the metamorphic core or, in stratigraphic terms, toward the basin margin. This usually results in an increase in the overall density of faulting in the same direction. 4. Structures become progressively younger in the direction of tectonic transport. 5. Rocks involved in thrusting also become younger in this direction. 6. Advancing thrust sheets act to load and depress the crust into local foreland basins (we know these as molasse basins). These fill with coarse marine and nonmarine detritus, which then also becomes involved in deformation. Such basins are often rich in plant-derived organic matter. 7. In many cases, a regional foreland basin, relatively rich in petroleum, will exist immediately in front of the thrust belt, presumably created by crustal loading on a more massive scale.
Convergent Basins: Non-Basement -Involved Faulting
The primary structural style in this category is the decollement foreland thrust-fold belt. Due to the thorough work of such authors as Price and Mountjoy (1970), Dahlstrom (1970), and Bally, Gordy and Steward (1966), the Canadian Cordillera has come to be generally treated as a type locality for such deformation. Figure 1 is the classic cross section by Price and Mountjoy (1970) through the southern portion of this structural belt and shows most of the principal features we have come to expect in this style.
Figure 1
The majority of hydrocarbon production exists in the foothills region, where faulting is especially complex. Thrust-faulted anticlines appear to be the most common traps in productive foreland regions throughout the world. But with regard to the size and specific geometry of traps, it should be emphasized that considerable variety exists within and between these regions. Over the span of Proterozoic time, differences in the size and morphology of plates, in their marginal sedimentary character, inherited structural features, and specific motions have all ensured a high degree of structural variation in foreland orogenic zones. Like compressive basement faulting, decollement thrusting is most often related to processes occurring along convergent plate boundaries, both in the mentioned foreland belts and along the leading edges of subduction zones. Figure 2 displays the interpreted structure on a seismic section across the active subduction complex offshore of northern California.
Figure 2
This section shows the de'collement faulting in the accretionary prism quite well. We note the change from the chaotically deformed accretionary prism sediments into the more conformable forearc basin at the far right in the figure. The deformed complex is broken into thrust slices about 400-500 m thick, whose bounding faults display variable curvature. This is probably due to a degree of stratigraphic control and ramping of fault development. We can also see that where compressional features dominate the accretionary prism, extensional faulting characterizes the deep sea sedimentary section west of the trench. This has been observed in a number of active subduction zones. Often, these faults have developed in the underlying ocean crust and have been interpreted as being the result of lithospheric bending and consequent tensional rupture. Collision-related thrust belts ( Figure 3 ) verge toward the subducting ("oncoming") plate.
Figure 3
A transition to basement-involved thrusting often occurs in these settings behind the thrust front. Figure 4 shows how the involvement is usually interpreted to increase into the mobilized, metamorphic "core" of an orogen, such as the Canadian Cordillera.
Figure 4
In the case of trench deformation, the situation is less well-understood. 'Two forms of basement-involved faulting are presumed-that which incorporates "slivers" of ocean crust, known as ophiolites into the accretionary prism, and that which cuts continental crust. For both trench and foreland thrust zones, overlap occurs between detached and "connected" structural styles. Most thrust belts presently exposed at the earth's surface occur as sinuous belts up to thousands of kilometers long. Their width is not uniform, but is instead characterized by sharply recessed and more gently extended portions known, respectively, as reentrants and salients. The cause for this type of variation along tectonic strike is not well-understood, but often thought to be related to preexisting basement features. In discussing foreland thrust belts, there are a number of general aspects that are of direct importance to hydrocarbon exploration: 1. Anticlines occur primarily in the hanging wall, and are asymmetric toward the direction of tectonic transport. 2. The geometry of thrusting is much dependent on the competence of the sedimentary units involved. The number of thrusts, as well as the intensity of folding, decreases in distinct proportion to increases in the amount of thick, competent units (e.g., sandstone, carbonates). 3. Sedimentary thicknesses decrease and deposits change character to more shallow-water clastics in the direction of tectonic transport, i.e., away from the metamorphic core or, in stratigraphic terms, toward the basin margin. This usually results in an increase in the overall density of faulting in the same direction. 4. Structures become progressively younger in the direction of tectonic transport. 5. Rocks involved in thrusting also become younger in this direction. 6. Advancing thrust sheets act to load and depress the crust into local foreland basins (we know these as molasse basins). These fill with coarse marine and nonmarine detritus, which then also becomes involved in deformation. Such basins are often rich in plant-derived organic matter. 7. In many cases, a regional foreland basin, relatively rich in petroleum, will exist immediately in front of the thrust belt, presumably created by crustal loading on a more massive scale.
Strike-Slip Basins
Strike-slip, or non-arc, basins are formed along convergent margins where the plates move by transcurrent faulting. Strike-slip basins are small basins that form through a combination of both the transcurrent fault movements and local block faulting. Strike-slip basins include the California basins, the Vienna Basin, and the Crimea and Baku Basins of the Soviet Union, and many others. Wrench Faults For most exploration purposes, strike-slip, oblique-slip, and wrench fault systems can be considered as roughly equivalent, though their specific plate tectonic settings are variable (see Appendix A). In all cases, they are assumed to involve basement. These fault systems can be pure strike-slip or they may include components of compression and extension, sometimes called transpression and transtension, respectively. Major wrench faults are usually associated with transform plate margins. In our general discussion of faulting, we have looked at several examples of these structures (we refer to Figure 1, Figure 2 , and Figure 3 ), which can occur as single shear planes or systems of parallel faults (Harding, 1974, 1983 and 1985).
Figure 1
Figure 2
Figure 3
For petroleum exploration purposes, we are most interested in the structures associated with wrench faults, since they form potential structural traps.
To a large extent, the components of compression and extension are determined by the degree of obliqueness involved in plate convergence. As convergence swings from "head-on" to lower and lower angles of incidence, the relative amount of strikeslip motion increases. Theoretically, associated compression should decrease. Wrench fault structures are extremely varied, as mentioned by Harding (1985), and include many of the features seen in other styles. Referring once again to Figure 3, we see how wrench fault structures can be resolved according to the complex stress fields that result from such large-scale shearing. The development of specific structures and their relative importance is naturally dependent on how much compression or extension might be involved in plate convergence. Flower Structures The term flower structure is often used to describe the upward-branching form that many wrench faults have been interpreted to show on seismic profiles. The flower itself is denoted positive or negative on the basis of whether the units within it are arched by reverse components of displacement or dropped by normal separations, respectively. In Figure 4 we see an interpreted example of a negative flower.
Figure 4
Note that seismic data is relatively chaotic immediately east of the interpreted fault. Letters "A" and "I" refer to the blocks whose relative motion is away from or toward the viewer.
Despite the relatively wide use of flower structure terminology, questions remain as to its value. Often, the quality of seismic data in the crucial core areas of supposed flowers is poor, which leads to inconclusive and alternate interpretations. Some structures that have been called positive flowers (e.g., in the Ardmore Basin of southern Oklahoma; see Harding and Lowell, 1979) have also been interpreted as thrusts. Those with apparent negative "bloom" have also been analyzed as rift structures. In regions such as the North Sea or the Red Sea, the problem is most likely compounded by the fact that components of both strike-slip and normal displacement are responsible for the fault geometries observed. Figure 6, Figure 7 , Figure 8 , Figure 9, Figure 10, and Figure 11 gives a range of structures from various styles that can appear similar to each other and to wrench fault structures in cases where data are poor and fault offset cannot be determined.
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
These examples highlight the potential problems that can arise from identifying any example of highly complex faulting as wrench-related.
Cratonic Basins There are two major types of basins that are located in the interior of plates on the stable continental crust or craton. The first cratonic basin type forms only in the central areas of continents, so these basins have often been called interior or intracratonic basins. We also sometimes call them sag basins because they form as simple sags or depressions on the underlying basement but have no adjacent highlands. In plan view, they are circular to elliptical in shape. In cross section, they appear saucer-like-symmetrical, shallow and somewhat flattened. The origin of interior basins is still poorly understood. Many of them are known to have block-faulted basement rock and attenuated crust, so the processes involved are mainly extensional. Possibly these basins may form in areas of aborted rifting, where dense material was injected into the lower crust, reducing crustal buoyancy (Dickinson, 1976).
Some of the largest sedimentary basins in the world belong to this class. Examples include the Williston, Illinois, and Michigan Basins of the central United States, and the Paris Basin of France. A second cratonic basin type, called foreland or craton margin basins, develops more toward the margins of continents. In a foreland basin, the first depositional cycle is usually interrupted by uplift. Then a second depositional cycle begins, most commonly from a different direction. A regional unconformity separates these two cycles. Although foreland basins are usually initiated by extension, the dominant processes in the second cycle are compressional, related to the uplift and orogeny. Because they evolve in two stages, foreland basins are sometimes called composite basins. They are usually elliptical or elongate, vary widely in size, and have asymmetric profiles. North America oil-producing foreland basins include the Alberta, Anadarko, Permian and Appalachian Basins, and most of the small Rocky Mountain basins. Vertical Faults Hydrocarbons are often trapped by secondary faults and closures within cratonic basins ( Figure 1 ).
Figure 1
The patterns of such smaller-scale structures are completely singular to each basin, but usually display pronounced trends that appear to be related to large-scale tectonic patterns in the underlying basement that have been identified from gravity and magnetic data. Generally speaking, then, basement structure is often an essential key to exploration of the potential in overlying sediments. Drilling, however, rarely penetrates the entire depth of sedimentary fill, and thus, direct study of basement rocks is usually limited to the shallower margins of a basin. Faults are most often high-angle and show either normal or strike-slip displacement. Compressional structures are usually open folds that may, as in the case of the
Michigan Basin ( Figure 2 ), be consistent in orientation.
Figure 2
Figure 1
The seismic section shown in Figure 1 shows the crestal collapse of a salt feature. Interpret the faults as indicated on this line. See the interpreted section. ( Figure 1
Figure 1
)
Figure 1
In the seismic section shown in Figure 1, we can see a reflection off the glide plane of the fault. Why is the glide plane reflection in this section a continuous event, rather than a discontinuous series of positive, zero and negative reflection coefficients?
Figure 6
What thrust styles, as defined in Figure 1 ,
Figure 7[a What thrust styles as defined in Figure 1
Figure 2 , Figure 3 , and Figure 4 , are exhibited in Figure 5 , Figure 6 , and Figure 7 ?
Figure 1
Figure 1 -- Style c Figure 2 -- Style a
Figure 2
Figure 3
Figure 3
-- Style b
Figure 1
The synthetic seismic section shown in Figure 1 was generated from a geological model. Sketch what you think the input model looks like.
Figure 1
Figure 1 .
Figure 1
What can you say about the timing of the thrust faults present in the seismic section shown in Figure 1 ? The thrusting occurred very recently, probably after the generation and migration of hydrocarbons. ( Figure 1
Figure 1
)
Figure 1
Is the flower structure depicted in the seismic section shown in Figure 1 positive or negative? Overall, the flower is a negative structure, as defined by the outer boundary faults, but it also has a positive component, as indicated by the inner fault. Figure 1
Figure 1
Figure 1
The seismic section shown in Figure 1 is from an offshore cratonic basin (the Sao Luiz Basin, offshore Brazil). Interpret the faults as indicated in this section.
Figure 1
Overview Workstations have given us new ways to manipulate and display seismic data. These new visualizations, in turn, have given us the ability to incorporate more information and detail into our interpretations, leading to higher success rates and more production. For example, we can now join multiple seismic lines at their intersections to form one composite display. This allows multiple wells to be tied together on one seismic display, even if no one, direct line exists between them. An example of a composite display tying two wells (even though no data were acquired directly between them) is shown in Figure 1.
Figure 1.
We notice in this example how the same fault can be observed on two of the lines. Another example of a workstation graphic is a concertina display. This display is a series of panels on the screen, which may show multiple parallel seismic lines, or it may be the same line but which different attributes such as frequency, phase, amplitude, etc. Sometimes this type of display is useful in picking very small faults. The input to the two examples addressed above is 2-D data. It is not a requirement to have 3-D data to conduct an interpretation on a workstation. However, 3-D data is superior to 2-D, particularly in defining faults and structure. A 3-D data survey has two substantial advantages over 2-D data. First, the vast amount of data in a 3-D survey allows us to see much greater detail than was possible using 2-D data. This detail has led to better definition of structure, direct hydrocarbon indicators and reservoir characteristics. And as we have already seen, this greater detail has altered our beliefs on the distribution and behavior of faults, as well as structural contours. The other advantage is the ability of the workstation to display the 3-D data in many new formats that are impossible to do with 2-D data. These new displays allow us to view the data in many new sections and dimensions. Data can now be displayed as a horizontal section, as a three-dimensional cube, in a combination of vertical and horizontal perspectives, as a random line extracted from the data that does not follow the orthogonal in line or cross line directions, or as sections of data that follow fault planes. And all of these graphics can be displayed from any azimuth. The result of these displays is maximum use of the data and better interpretations. Of all the new visualizations, the horizontal section is by far the most intriguing and informative. In this display, the data set is transsected horizontally along a defined time or depth and usually indicated by variable color density. The horizontal section is also called a slice if it along an isotime, it is called a time slice; if it is along a particular depth, it is called a depth slice. Figure 2 ,
Figure 2
Figure 3, and Figure 4 show time slices from the Gulf of Mexico, the Gulf of Thailand and onshore Europe respectively.
Figure 3
Can you see the faults in these displays?
Figure 4
With the capability to create a time slice, several other displays are now possible. A chair or folded section display is a composite of the conventional vertical section and a horizontal time slice. It is obvious in Figure 5 why this type of graphic is often called a "chair" display.
Figure 5
We notice how the fault is presented in each of the sections and how it "ties" at the corners. Figure 6 shows another display created with a time slice.
Figure 6
This is called a cube display. Again we notice how the fault is presented in this threedimensional display and how we can follow it in three directions. In all the example displays shown above, we see how the faults can be easily followed around corners, from line to line, and in a horizontal perspective. Faults can be more accurately interpreted and mapped using such displays. Another display that aids in the identification of faults is the dip map. A dip map is a projection of an interpreted horizon such that the dip magnitude (throw) and/or dip direction (angle) is calculated and displayed. Figure 7 is an example of a dip map.
Figure 7
In this map, colors are used to indicate dip magnitude. Any linear contact that shows an abrupt change in color, such as we see here, indicates the presence of faults. The more pronounced the color change, the larger the dip. Figure 8 and Figure 9 are two more examples of dip maps.
Figure 8
In this case, both maps are of the same area, the Nun River Field in Nigeria.
Figure 9
Figure 8, indicates the dip magnitude, while Figure 9 indicates dip azimuth. Again the faults are identifiable by their linear appearance and abrupt change in color. We notice the ease of identifying faults in these displays and the wealth of information that can be obtained. In particular, we notice how the faults appear to be discontinuous along the same strike line.
Time Slices Interpreting faults in horizontal time slices is very similar to that in traditional vertical sections. As in seismic sections, faults in time slices are discernible by a termination and/or displacement of aligned events. Let us look at the time slice in Figure 1 .
Figure 1
As an exercise, try picking the faults on this time slice. The event terminations indicating the presence of faults are clearly visible in this particular display. Of course, using vertical sections in conjunction with horizontal sections makes fault identification and interpretation easier. Figure 2 shows one interpretation of Figure 1 .
Figure 2
Viewing consecutive time slices in rapid succession on a workstation can aid in the identification of faults. This type of viewing is called a movie. In a time slice movie, horizon reflections appear to move across the screen, while faults generally remain stationary or move only slightly. This visual effect is due to the vertical nature of fault planes relative to the seismic reflections. The higher the angle of the fault, the more stationary the fault will appear; low-angle faults will move slowly across the screen. Another characteristic of a time slice movie is the ability to see horizons move across the screen and then truncate against the faults. Using this technique, we are able to recognize and locate smaller faults and be more precise with their fault boundaries. Figure 3 ,
Figure 3
Figure 4 ,
Figure 4
Figure 5 , and Figure 6 show four frames of a time slice movie.
Figure 5
Notice the movement of the horizon reflections versus the stationary appearance of the faults.
Figure 6
Sometimes interpreting faults in time slices is not easy. This is especially true when the trend, or strike, of the structural features and contours is the same as the strike of the fault. When this occurs, the event terminations are difficult to see, and therefore, the faults are likewise difficult to identify. The more parallel the fault strike is to the structural strike, the more difficult it is to see the event terminations. Conversely, the more the faults are in the dip direction, the easier it is to see the event terminations. Let us look at the time slice from the Gulf of Thailand in Figure 7 .
Figure 7
Some of the major faults in this time slice are easy to recognize while others are nearly impossible. This is because the structural and fault strike are generally parallel. Again, as an exercise, try picking the faults on this time slice. Let us compare Figure 7 to Figure 8 .
Figure 8
Figure 8 indicates a fault interpretation of the time slice data in Figure 7 . The interpretation is based on both horizontal and vertical sections. In this example, if the interpretation was based on the time slice alone, many of the faults might have been overlooked. Time slices can also be used to generate various fault plane projections. Displaying faults interpreted on sequential time slices in a 3-D format provides a simple means to do so. Figure 9 shows such an example.
Figure 9
Here, the fault planes are displayed in a cubic projection. Essentially, this type of display is a three-dimensional view of fault plane maps, but better, because it gives us the ability to validate a fault interpretation in three dimensions.
Fault Slices One of the more innovative displays of 3-D data and the workstation is the fault slice. A fault slice ( Figure 1 ) is a projection of data along a surface parallel to an interpreted fault plane. It is displayed as a vertical section.
Figure 1
We use a fault slice for determining fault throw, sealing and leaking properties, juxtaposition of lithologies, planning of deviated wells, identifying splinter faults, determining other fault related criteria, or simply mapping detailed structure along the fault plane. To generate a fault slice, we begin by identifying a fault in the 3-D data set. The fault of interest must be interpreted at several horizons in order for the computer to calculate the fault plane's projection in x, y, t space as in Figure 2 .
Figure 2
The more the fault plane deviates from a straight line in the x, y or t directions, the more data points are required for an accurate projection. Once the fault plane data points are defined, the computer will then fit a linear projection or slice to the data points by applying a triangulation, least squares or other statistical method of our choice. Any data measurement, such as amplitude, frequency, acoustic impedance, or throw, can be plotted in the fault slice. The type of data displayed is dependent on the purpose of the investigation. For structural detail, we could display the throw as in Figure 3.
Figure 3.
We could also display amplitudes to define the splinter faults as in Figure 4.
Figure 4.
Case Studies 3-D data have altered our knowledge and understanding of faults. Where we once thought a few long, continuous faults existed, we now know many short ones to exist instead. This new understanding is a direct result of the increased data coverage associated with 3-D survey. For example, previously, a tight 2-D grid was once considered to be 2 km x 2 km (2,000 m x 2,000 m). Now, with 3-D surveys, a normal grid is 25 m x 25 m! Instead of 4 data points per 4 sq km, we now have 6,400! This increase in data points has provided us with new detail, particularly in the number of faults, their length and the shape of structural contours. To exemplify this observation, we will look at three case studies comparing interpretations based on a 2-D seismic survey to that based on a 3-D survey. The first case study (Brown, 1991) is of the Gulf of Thailand area and data, shown in Figure 1 and Figure 2 . Figure 3 shows the structure map derived from a 2-D grid
while Figure 4 shows the structure map of the same area derived from a more recent 3-D survey.
Figure 1
Figure 2
Figure 3
Figure 4
Notice the remarkable change in the fault interpretation. Whereas the 2-D interpretation indicates three fault systems with a series of intermingling faults, the 3-D interpretation indicates many more separate and abbreviated faults. Also when we compare the differences in the contour detail, we notice how the contours in the 3-D interpretation along the faults are very much parallel to the fault strike. Despite their equal contour intervals, we observe much more detail in the map based on the 3-D survey. Similar results for several prospects and fields were reported by Nestvold (1992).
Figure 5
Figure 5, which is from that study, offers another comparison of a fault pattern based on 2-D data versus another of the same area but based on 3-D data. Again, we notice how the faults have increased in number and have shortened in length. In the some cases, even the orientation of the faults changes. In another example reported by Nestvold (1992), a map of the Groningen Gas Field based on 2-D data indicates the principal faults to trend in a northwest/southeast direction. The more recent interpretation, based on 3-D seismic, indicates that the dominant faulting actually trends north/south with secondary east/west cross faulting. These maps are shown in Figure 6 and Figure 7 .
Figure 6
Figure 7
Another case study showing the influence of 3-D data was reported by Bouvier et al. (1989). In this study of the Nun River Field in Nigeria, the 3-D data not only altered the structural interpretation, but also provided a means to investigate the sealing and leaking properties of the major faults. Two vintages of Nun River structure maps are shown in Figure 8 and Figure 9 .
Figure 8
Figure 9
The first map ( Figure 8 ), based on 2-D data, shows a collapsed anticlinal crest dominated by long, continuous normal faults. The second map ( Figure 9 ), based on 3-D data, shows that the major faults have the same strike as previously mapped, but the complexity and discontinuity of the faulting have increased. The major faults in the 3-D interpretation are discontinuous and rapidly vary in throw and hade. In addition, many minor faults that had been previously unidentified were recognized and mapped. In this same Nun River Field study by Bouvier et al. (1989), fault slices were used to determine the sealing and leaking characteristics of a major fault. In this study, fault slices of acoustic impedance in the depth domain were generated in order to identify lithologies and clay smears along the fault plane. Figure 10 shows the acoustic impedance fault slice for the footwall.
Figure 10
The lithologies and fluids along the footwall fault plane were determined by combining this information with well logs. This procedure was also followed for the hanging wall. By superimposing the interpreted fault slices from the hanging and footwalls, the juxtaposition of impermeable clays and shales against hydrocarbonbearing sandstones was determined, as was a measure of fault sealing and leaking.
References
Allen, C. R. 1962. Circum-Pacific faulting in the Philippines-Taiwan region. Jour. Geophys. Res. 67: 4795812. Anstey, N. A. 1992. GP204 The Geological Message in the Seismic Trace: Boston: IHRDC. Anstey, N. A. 1990. GP203 The reflection process. Boston: IHRDC. Anstey, N. A. 1987. GP201 Wiggles: A graphical introduction to signal theory. Boston: IHRDC.
Anstey, N. A. 1983. GP102 The seismic method. Boston: IHRDC. Anstey, N. A. 1983. GP101 The exploration task. Boston: IHRDC. Anstey, N. A. and M. E. Badley. 1984. GP501 Basic interpretation. Boston: IHRDC. Anstey, N. A. and M. E. Badley. 1988. GP502 Seismic contouring. Boston: IHRDC. Arthur, K. R., D. R. Cole, G. L. Henderson and D. W. Kushnir. 1983. Geology of the Hibernia discovery. AAPG Memoir no.32:181-197. Atwater, T. 1970. Implications of plate tectonics for the Cenozoic tectonic evolution of western North America. GSA Bull. 81: 3513-3536. Badley, M. E. 1985. Practical seismic interpretation. Boston: IHRDC. Bally, A. W., ed. 1983. Seismic expression of structural styles: A picture and work atlas. AAPG Studies in Geology Series, no. 15 (3 vols.). Tulsa, OK: AAPG. Bally, A. W., P. L. Gordy and G. A. Steward. 1966. Structure, seismic data, and orogenic evolution of southern Canadian Rocky Mountains. Can. Petrol. Geol. Bull. 14: 337-381. Bally, A. W. and J. S. Oldow. 1983. Plate tectonics, structural styles, and the evolution of sedimentary basins. Houston Geological Society, short course manual. Barbat, W. F. 1958. The Los Angeles Basin area, California. Habitat of oil, L. G. Weeks, ed.: 62-77. Tulsa, OK: AAPG. Biddle, K. T. and D. R. Seely. 1983. Structure of a subduction complex. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 3.4.2-129-135. Billings, M. P. 1972. Structural geology (third edition). Englewood Cliffs, NJ: Prentice-Hall. Bishop, R. S. 1978. Mechanism for emplacement of piercement diapirs. MPG Bull. 62:1561-1583. Bloom, A. L. 1978. Geomorphology. Englewood Cliffs, NJ: Prentice-Hall. Bobbit, J. E. and J. 0. Gallagher. 1978. The petroleum geology of the Gulf of Suez. 1978 Offshore Tech. Conf, Houston, TX. 37-382. Bouvier, J. D., C. H. Kaars-Sijpesteijn, D. K Kluesner, C. C. Onyejekwe and R. C. Van Der Pal. 1989. Three-dimensional seismic interpretation and fault sealing investigations, Nun River Field, Nigeria. MPG Bull. 73: 1397-1413. Brink, A. H. 1974. Petroleum geology of the Gabon Basin. MPG Bull. 58: 216-235.
Brown, A. R. 1991. Interpretation of three-dimensional seismic data (third edition).MPG Memoir 42. Tulsa, OK: AAPG. Brown, L. F., Jr. and W. L. Fisher. 1977. Seismic-stratigraphic interpretation of depositional systems: examples from Brazilian rift and pull-apart basins. Seismic stratigraphy-Applications to hydrocarbon exploration. AAPG Memoir no. 26, C. E. Payton, ed. 213-248. Bruce, C. H. 1973. Pressure shale and related sediment deformation-Mechanisms for development of regional contemporaneous faults. MPG Bull. 57: 878-886. Bruce, C. H. 1983. Shale tectonics, Texas coastal area growth faults. Seismic Expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 3.1-1-6. Bullard, T. 1973. Eugene Island Block 126 Field, offshore St. Mary Parish, Louisiana. Offshore Louisiana oil and gas fields. Lafayette Geol. Soc.: 111. Casella, C. J. 1964. Geologic evolution of the Beartooth Mountains, Montana and Wyoming-Part 4: Relationship between Precambrian and Laramide structures in the Line Creek area. GSA Bull. 75: 969-986. Chapin, C. E. and S. M. Cather. 1983. Eocene tectonics and sedimentationColorado Plateau, Rocky Mountain area. Rocky Mountain foreland basins and uplifts, J. D. Lowell, ed. RMAG: 33-57. Christensen, C. K 1983. An example of a major syndepositional listric fault. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 2.3.1-30. Cloos, E. 1947. Oolite deformation in the South Mountain Fold, Maryland. GSA Bull. 58: 843-918. Cloos, H. 1939. Conversations with the Earth: Munich: Piper & Co. Cook, F. 1982. The COCORP seismic reflection traverse across the southern Appalachians. AAPG Studies in Geology Series no. 14: 61p. Cook, F. et al. 1979. Thin-skinned tectonics in the crystalline southern Appalachians; COCORP seismic reflection profiling of the Blue Ridge and Piedmont. Geology 7: 563-567. Crowell, J. C. 1974. Origin of late Cenozoic basins in southern California. Tectonics and Sedimentation. SEPM Special Publications no. 22: 190-204. Currelle, R. and R. Marco. 1983. Reflection profiles across the Aquitaine Basin (salt tectonics). Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 2.3.2-11-18. Dahlstrom, C. D. A. 1970. Structural geology in the eastern margin of the Canadian Rocky Mountains. CSPG. Bull. 18: 33206.
Dailly, G. C. 1976. A possible mechanism relating progradation, growth faulting, clay diapirism, and over-thrusting in a regressive sequence of sediments. CSPG Bull. 24: 92-116. Davies, P. A. and S. A. Runcorn. 1980. Mechanics of continental drift and plate tectonics. London: Academic Press. Davis, G. H. 1984. Structural geology of rocks and regions. New York: John Wiley & Sons. De Jong, K. A. and R. Scholten, ed. 1973. Gravity and tectonics. New York: John Wiley & Sons. Dennis, J. G. 1972. Structural geology. New York: John Wiley & Sons. De Sitter, L. U. 1964. Structural geology. New York: McGraw-Hill. Dickinson, W. R. 1976. Plate tectonic evolution of sedimentary basins. Plate tectonics and hydrocarbon accumulation. AAPG Course Note Series no.1, 1-56. Dieterich, J, H. 1969. Origin of cleavage in folded rocks. Am. Jour. Sci. 267: 155165. Dillion, W. P., C. K. Paull, R. T. Buffler and J. P. Fail. 1979. Structure and development of the southeast Georgia embayment and northern Blake plateau: Preliminary analysis. Geological and geophysical investigations of continental margins. AAPG Memoir no. 29, J. S. Watkins, L. Montadert and P. W. Dickerson, ed. 273. Douglas, R. J. W. 1950. Callum Creek, Langford Creek, and Gap map-areas, Alberta. Geol. Survey of Canada Memoir no.255. Elliot, D. 1976. The motion of thrust sheets. Jour Geophys. Res. 81: 949-963. Erxleben, A. W., and G. Carnahan. 1983. Slick Ranch area, Starr County, Texas. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 2.3.1-22-27. Evans, R. 1978. Origin and significance of evaporites in basins around the Atlantic margin. AAPG Bull. 62: 223-234. Fischer, A. G. 1969. Geologic time-distance rates: the Bubnoff unit. GSA Bull. 80: 549-551. Fischer, A. G. and S. Judson, ed. 1975. Petroleum and global tectonics. Princeton, NJ: Princeton Univ. Press. Flint, R. F. 1971. Glacial and quaternary geology. New York: John Wiley & Sons.
Fox, K G. 1959. Structure and accumulation of hydrocarbons in southern foothills, Alberta, Canada. AAPG Bull. 43: 992-1023. Gansser, A. 1964. Geology of the Himalayas: New York: Interscience. Gibbs, A. D. 1983. Balanced cross-section construction from seismic sections in areas of extension tectonics. Jour. Struct. Geol. 5 (2)153. Gibson, I. L. and G. P. L. Walker. 1964. Some composite rhyolite-basalt lavas and related composite dykes in eastern Iceland. Proc. Geol. Assoc. London 74: 301-318. Gries, R. 1983. North-south compression of Rocky Mountain foreland structures. Rocky Mountain Foreland Basins and Uplifts (J. D. Lowell, ed.): RMAG: 9-33. Griggs, D. T., and J. Handin. 1960. Observations on fracture and a hypothesis of earthquakes. Rock Deformation - A Symposium, GSA Memoir no. 79 (D. T. Griggs and J. Handin, ed.): 347-373. Grow, J. A., R. E. Mattick, and J. S. Schlee. 1979. Multichannel seismic depth sections and interval velocities over outer continental shelf and upper continental slope between Cape Hatteras and Cape Cod. Halbouty, M. T. 1979. Salt domes, Gulf Region, United States and Mexico, 2nd ed.: Houston: Gulf Publ. Co. Handin, J. W., R. V. Hager, M. Friedman, and J. M. Feather. 1963. Experimental deformation of sedimentary rocks under confining pressure: pore pressure tests. AAPG Bull. 47: 717-755. Harding, T. P. 1985. Seismic characteristics and identification of negative flower structures, positive flower structures, and positive structural inversion. AAPG Bull. 69: 582-601. Harding, T. P. 1984. Graben hydrocarbon occurrences and structural style. AAPG Bull. 68: 333-362. Harding, T. P. 1983a. Divergent wrench fault and negative flower structure, Andaman Sea. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 4.2-1-9. Harding, T. P. 1983b. Structural inversion at Rambutan Oil Field, South Sumatra Basin. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 3.3-13-18. Harding, T. P. 1976. Tectonic significance and hydrocarbon trapping consequences of sequential folding synchronous with San Andreas faulting, San Joaquin Valley, California. AAPG Bull. 60: 356-378. Harding, T. P. 1974. Petroleum traps associated with wrench faults. AAPG Bull. 58: 1290-1304.
Harding, T. P. and J. D. Lowell. 1979. Structural styles, their plate-tectonic habitats and hydrocarbon traps in petroleum provinces. AAPG Bull. 63: 1016-1069. Hatcher, R. D., Jr. 1972. Development model of the southern Appalachians. GSA Bull. 83:2735. Heard, H. C. 1960. Transition from brittle to ductile flow in Solenhofen limestone as a function of temperature, confining pressure, and interstitial fluid pressure. Rock deformation-A symposium. GSA Memoir no. 79, D. T. Griggs and J. Handin, ed. 193226. Heard, H. C. and C. B. Raleigh. 1972. Steady-state flow in marble at 500 C to 800 C. GSA Bull. 83: 935-956. Hills, E. S. 1971. Elements of structural geology (second edition). London: Methuen and Co., Ltd. Hobbs, B. E., W. D. Means and P. E. Williams. 1976. An outline of structural geology. New York: John Wiley & Sons. Hoeppener, R. 1964. Zur physikalischen tektonik. Felsmechanik und Ingenjeurgeologie 2: 224. Holland, D. S., C. E. Sutley, R. E. Berlitz and J. A. Gilreath. 1976. East Cameron Block 270: A Pleistocene field. North American oil and gas fields. AAPG Memoir no. 24, J. Braunstein, ed. 205-228. Huff, K. F. 1980. Frontiers of world exploration. Facts and principles of world petroleum occurrence. CSPG Memoir no. 6, A. D. Miall, ed. 343-362. Huff, K. F. 1978. Frontiers of world exploration. Oil and Gas Jour. 76:214-220. Inoue, E. 1960. Land deformation in Japan. Geograph. Sun;. Inst (Japan) Bull. 6: 73-134. Jackson, J. and D. P. McKenzie. 1983. The geometric evolution of normal fault systems. Jour. of Structural Geology 5: 47182. Jackson, M. P. and W. E. Galloway. 1984. Structural and depositional styles of Gulf Coast tertiary continental margins: Application to hydrocarbon exploration. AAPG Continuing Education Course Notes Series, no. 25. Jenyon, M. K. and A. A. Fitch. 198 5. Seismic reflection interpretation. Berlin: Gebruder Bomtraeger. Kennedy, W. G. 1946. The Great Glen Fault. Quart. Jour. Geol. Soc. London 102: 41-72. Klemme, H. D. 1976. World oil and gas reserves from analysis of giant oil fields and petroleum basins (provinces). The future supply of nature-made petroleum and gas, R. K Meyer, ed. New York: Pergamon Press: 217-260.
Kleyn, A. H. 1983. Seismic reflection interpretation. London: Applied Science Publishers. Lamerson, P. R. 1982. The fossil basin area and its relationship to the Absaroka thrust fault system. Geologic studies of the Cordilleran thrust belt, R. B. Powers, ed. RMAG: 279-340. Lehner, P., H. Doust, G. Bakker, P. Allenbach and J. Gueneau. 1983. Active margins-Caribbean margin of South America, profiles C-1422, C- 1412, and C-1413. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 3.4.2-111-129. Ligget, M. A. and H. E. Ehrenspeck. 1974. Pahranagat shear system, Lincoln County, Nevada. Argus Explor. Co., Rept. of Inv., NASA CR 1356388, E74-10206. Lowell, J. D. 1985. Structural styles and petroleum exploration. Tulsa, OK: OGCI Pub. Lowell, J. D., G. J. Genik, T. H. Nelson and P. M. Tucker. 1975. Petroleum and plate tectonics of the southern Red Sea. Petroleum and global tectonics, A. G. Fisher and S. Judson, ed. Princeton, NJ: Princeton Univ. Press. 129-153. Maher, C. E. 1980. Piper Oil Field. Giant oil and gas fields of the decade 1968-1978. AAPG Memoir no. 30, M. Halbouty, ed. 131-173. Manspeizer, W. 1981. Early Mesozoic basins of the central Atlantic passive margins. AAPG Continuing Education Course Notes Series, no. 19. Martin, R. G. and J. E. Case. 1975. Geophysical studies in the Gulf of Mexico. The Oceans Basins and Margins, V. 3, A. E. Nairn and F. G. Stelhi, ed. New York: Plenum Press. Mescherikov, Y. A. 1968. Neotectonics. Encyclopedia of geomorphology. R. W. Fairbridge, ed. New York: Van Nostrand. 768-773. Mitchum, R. M., P. R. Vail and S. Thompson. 1977. The depositional sequence as a basic unit for stratigraphic analysis. Seismic stratigraphy-Applications to hydrocarbon exploration. AAPG Memoir no. 26, C. E. Payton, ed. 53-62. Montgomery, S. L. 1984. Michigan Basin: Expanding the deep frontier. Petroleum Frontiers 1 (Spring). Denver: Petroleum Information Corp. Montgomery, S. L 1987. GP601 Structural geology. Boston: IHRDC. Morgan, L. and W. Dowdall. 1983. The Atlantic continental margin. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 2.2.3-30-36. Nadai, A. 1950. Theory of flow and fracture of solids. New York: McGraw-Hill. Nettleton, L. L. 1934. Fluid mechanics of salt domes. AAPG Bull. 18:1175-1204.
Nestvold, E. 0. 1992. 3-D seismic: Is the promise fulfilled? The Leading Edge (SEG) 11(6): 12-19. New Orleans Geological Society. 1983. Oil & gas fields of southeast Louisiana, v. 3. New Orleans, LA. Park, R. G. 1983. Foundations of structural geology. London: Chapman & Hall. Pew, E. 1983. Seismic structural analysis of deformation in the southern Mexican Ridges. Master's thesis: University of Texas, Austin, TX. Pierce, W. G. 1957. Heart Mountain and South Fork detachment thrusts of Wyoming. AAPG Bull. 41: 591-626. Pieri, M. 1983. Three seismic profiles through the Po Plain. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 3.4.18-26. Plafker, G. 1965. Tectonic deformation associated with the 1964 Alaskan earthquake. Science 148:1675-1687. Price, R. A. and E. W. Mountjoy. 1970. Geologic structure of the Canadian Rocky Mountains between Bow and Athabasca Rivers-A progress report. Geol. Assoc. Canada Spec. Paper 6. 25 p. Ragan, D. M. 1984. Structural geology: An introduction to geometrical techniques (third edition). New York: John Wiley & Sons. Ramsay, J. G. 1967. Folding and fracturing of rocks. New York: McGraw-Hill. Ramsay, J. G. and M. I. Huber. 1985. The techniques of modern structural geology. London: Academic Press. Reading, H. G. 1980. Characteristics and recognition of strike-slip fault systems. Sedimentation in oblique-slip mobile zones. Int. Assoc. of Sed. Spec. Pub. 4, P. F. Ballance and H. G. Reading, ed. 7-26. Richter, C. K 1952. Elementary seismology. New York: W. H. Freeman. Robson, D. A. 1971. The structure of the Gulf of Suez rift, with special reference to the eastern side. Jour. Geol. Soc. 127: 247-276. Roeder, D. H. and 0. E. Gilbert. 1977. Structure, kinematics, and hydrocarbon prospects of thrust and fold belts. AAPG Structural Geology School Course Notes. Rodgers, J. 1953. The folds and faults of the Appalachian Valley and Ridge province. Kentucky Geol. Survey Spec. Pub. 1:150-166. Roux, W. K, Jr. 197 7. The development of growth fault structures. AAPG Structural Geology School Course Notes.
Royse, K, Jr., M. A. Warner and D. L. Reese. 1975. Thrust belt structural geometry and related stratigraphic problems, Wyoming-Idaho-northern Utah. Deep drilling frontiers in the central Rocky Mountains, D. W. Boylyard, ed. RMAG Symposium: 415. Sacrison, W. R. 1978. Seismic interpretation of basement block faults and associated deformation. Laramide folding associated with basement block faulting in the western United States, V. Mathews, ed. GSA Memoir no. 26. 165-184. Sbar, M. L. and L. R. Sykes. 1973. Contemporary compressive stress and seismicity in eastern North America: An example of intra-plate tectonics. GSA Bull. 84:1861-1881. Seely, D. R., and W. R. Dickinson. 1977. Structure and stratigraphy of forearc regions. AAPG Continuing Education Course Notes Series, no. 5. Selley, R. C. and D. C. Morrill. 1983. GL107 The habitat of hydrocarbons in sedimentary basins. Boston: IHRDC. Seni, S. J. and M. P. Jackson. 1984. Evolution of salt structures, east Texas diapir province, Parts 1 and 2. AAPG Bull. 67:1245-1274. Seyfert, C. K. and L. A. Sirkin. 1973. Earth history and plate tectonics. New York: Harper & Row. Sodbinow, E. S. 1984. GP401 Basic processing. Boston: IHRDC. Spicher, A. 1980. Tectonics profile through Switzerland. Geology of Switzerland, Swiss Geol. Comm. Spillers, J. P. 1965. Distribution of hydrocarbons in South Louisiana by type of traps and trends-Frio and younger sediments. Trans. Gulf Coast Assn. Geol Soc 15:37-39. Sprague, E. L. 1983. Geology of the Tepee Flats-Bulldog Fields, Natrona County, Wyoming. Rocky Mountain foreland basins and uplifts, J. D. Lowell, ed. RMAG: 339345. Stone, D. S. 1983a. Seismic profile-South Elk Basin. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 3: 3.4.18-26. Stone, D. S. 1983b. The Greybull sandstone pool (lower Cretaceous) on the Elk Basin thrust-fold complex, Wyoming and Montana. Rocky Mountain foreland basins and uplifts, J. D. Lowell, ed. RMAG: 345-357. Stude, G. R. 1978. Depositional environments of the Gulf of Mexico, South Timbealer Block 54: Salt dome and salt-dome growth models. Trans Gulf Coast Assn of Geol. Soc. 28: 627-644.
Sunwall, N. T., K. A. McQuillan and C. J. Nick. 1983. Salt diapir-Gulf of Mexico. Seismic expression of structural styles, A. W. Bally, ed. AAPG Studies in Geology Series no. 15, v. 2: 2.3.2-32-38. Swann, D. H. and A. H. Bell. 1958. Habitat of oil in the Illinois Basin. Habitat of oil, L. G. Weeks, ed. Tulsa, OK: AAPG. 44772. Tearpock, D. J. and R. E. Bischke. 19 91. Applied subsurface geological mapping. Englewood Cliffs, NJ: Prentice-Hall. Trusheim, F. 196 0. Mechanism of salt migration in northern Germany. MPG Bull. 44:1519-1541. Turcotte, D. and G. Schubert. 1982. Geodynamics. New York: John Wiley & Sons. Vail, P. K., R. N. Mitchum, R. G. Todd, J. N. Widmier, S. Thompson, J. N. Bubb and W. G. Hatfield. 1977. Seismic stratigraphy and global changes in sea level, Parts 1 and 2. AAPG Memoir no. 26: 49-213. Vine, K J. 1966. Spreading of the ocean floor: New evidence. Science 154: 14051415. Wallace, R. E. 1970. Earthquake recurrence intervals on the San Andreas Fault. GSA Bull. 81: 2875-2890. Watson, J. M. and C. A. Swanson. 197 5. North Sea-Major petroleum province. AAPG Bull. 59:1098-1112. Wennekers, J. H. L. 1958. South Sumatra basinal area. Habitat of oil, L. G. Weeks, ed. Tulsa, OK: AAPG. 1347-1358. Wilson, J. T. 1965. A new class of faults and their bearing on continental drift. Nature 207: 343-347. Withjack, N. 0. and C. Scheiner. 1982. Fault patterns associated with domes-An experimental and analytical study. MPG Bull. 66: 302-316. Wood, D. S. 1974. Current views of the development of slaty cleavage. Ann. Rev. Earth Sci. 2:1-35. Woodbury, H. 0., I. B. Murray, Jr. and R. E. Osborne. 1980. Diapirs and their relation to hydrocarbon accumulation. Facts and principles of world petroleum occurrence. CSPG Memoir no.6, A. D. Miall, ed. 119-143. Ziegler, P. A. 1980. Hydrocarbon provinces of the northwest European basins. Facts and principles of world petroleum occurrence. CSPG Memoir no.6, A. D. Miall, ed. 653-705
Additional References
Alam, A. and Caragounis, P. (1994). "Advances in 3D seismic fault interpretation." Abstract 56 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Anderson, J., Thomas, C., Kat, H. and Lynn, W. (1994). "Velocity and structural interpretation with 3D pre-stack depth migration." Abstract 56 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Anfiloff, V. (1989). "Discussion on 'Structural interpretation of the Rukwa Rift, Tanzania', by J. W. Peirce and L. Lipkov." Geophysics, 54, 1499-1500. Tulsa: Society of Exploration Geophysicists. Bally, A. W. (2001). "On the Role of Plate Tectonics in Hydrocarbon Exploration." Annual AAPG Meeting Expanded Abstracts, 11. Tulsa: American Association of Petroleum. Bally, A. W. (1991)."What Else to Do with Reflection Seismology." Abstract 61 st Annual International Meeting, 722 . Tulsa: Society of Exploration Geophysicists. Barton, C. A., Moos, D. and Zoback, M. D. (1997). "In-situ Stress Measurements Can Help Define Local Variations in Fracture Hydraulic Conductivity at Shallow Depth." The Leading Edge, 16, 11, 1653-1656 . Tulsa: Society of Exploration Geophysicists. Brown, A. R. (1983). "Structural Interpretation from Horizontal Seismic Sections." Geophysics, 48, 1179-1194. Tulsa: Society of Exploration Geophysicists. Brown, A. (1982). "Structural Interpretation from Horizontal Seismic Sections." Expand Abstract 52 nd Annual International Meeting. Tulsa: Society of Exploration Geophysicists. Bruno, M. S. and Winterstein, D. F. (1994). "Some Influences of Stratigraphy and Structure on Reservoir Stress Orientation." Geophysics, 59, 924-962. Tulsa: Society of Exploration Geophysicists. Chopra, S. and Sudhakar, V. (2000)." Fault Interpretation; the Coherence Cube and Beyond." Oil and Gas Journal,98, 31, 71-74. Tulsa: PennWell. Cloke, I. R., Craig, J. and Blundell, O. (1998). "Variations in Stress Regime through Time - their Effect on the Formation and Modification of the Kutai Basin, Kalimantan." Abstract 60 th Annual Meeting. Houten: European Association of Geoscientists & Engineers.
Colletta, B., Piron, E., Marchal, D. and Moretti, I. (2002). "Gravitational Folds and Compressional Structures Related with Normal Faults in the Suez Rift." Abstract 64 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Csato, I. and Foldes, T. (1999). "3D Visualisation of Architectural Patterns in Saltrift Basins." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers. Dancer, P.N. (2002). "The Corrib Field - Changes in Structural Interpretation through 2D, 3D and PreSDM Seismic Data." Abstract 64 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Dancer, P.N. (2001). "The Corrib Field - Impact of 2D, 3D and PreSDM Seismic Data on the Structural Interpretation." Abstract 63 rd Annual Meeting. Houten: European Association of Geoscientists & Engineers. Davies, R. K., Crawford, M., Dula, W. F., Jr., Cole, M. J. and Dorn, G. A. (1997). "Outcrop Interpretation of Seismic-scale Normal Faults in Southern Oregon: Description of Structural Styles and Evaluation of Subsurface Intepretation Methods" The Leading Edge, 16, 08, 1135-1141. Tulsa: Society of Exploration Geophysicists. Destro, N., Szatmari, P., Alkmim, F. and Magnavita, L. P. (2000). "Release Faults, Associated Structures, and their Control on Petroleum Trends in the Reconcavo Rift, Northeast Brazil." AAPG Bulletin, 87, 7, 1123-1144. Tulsa: American Asotiation Petroleum Geologists. Deville, E. and Collaku, A. (1999). "The Front of the Central Albanian Thrust Belt: Structural Interpretation and Sub-thrust Plays." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers. Driggs, A. F. (1997). "The Use of 3-D fault Visualization and Reservoir Juxtaposition Analysis in an Exploration Setting, U. S. Gulf of Mexico." Abstract American Association of Petroleum Geologists 1997 Annual Convention, 6, 70. Tulsa: American Association of Petroleum Geologists. Edwards, A. (1989). "Basin Margin Structural Styles, Jeanne d'Arc Basin Offshore Newfoundland, Canada." Abstract 59 th Annual International Meeting, 538 . Tulsa: Society of Exploration Geophysicists. Fehmers, G. C. and Hocker, C. (2003). "Fast Structural Interpretation with Structure-oriented Filtering." Geophysics, 68, 1286-1293. Tulsa: Society of Exploration Geophysicists. Gawthorpe, R., Jackson, C., Young M., Sharp, I., Moustafa, A. and Leppard, C. (2003). "Normal Fault Growth, Displacement Localization and the Evolution of Normal Fault Populations: the Hammam Faraun Fault Block, Suez Rift, Egypt." Journal of Structural Geology, 25, 6, 883-885. London: Reed Elsevier. Geraghty, D., Shannon, P. M and Feely, M. (1999). "Sedimentary Provenance of Rift Sediments of the Porcupine Basin, West of Ireland." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers
Gibson, R. I. (1994). "Structural Styles of the West Siberian Rift System and Their Impact on Hydrocarbon Accumulation." Abstract 64 th Annual International Meeting, 1626 . Tulsa: Society of Exploration Geophysicists. Goulty, N. R. (1998). "Relationships Between Porosity and Effective Stress in Shales." First Break, 16, 2, 413-419. Houten: European Association of Geoscientists & Engineers. Grandi, S., Rao, R., Huang, X. and Toksoz, N. (2003). "In Situ Stress Modeling at a Borehole - A Case Study." Abstract 73 rd Annual International Meeting, 297-300 . Tulsa: Society of Exploration Geophysicists. Gran t, J. and Kattenhorn, S. (2004). "Evolution of Vertical Faults at an Extensional Plate Boundary, Southwest Iceland." Journal of Structural Geology, 26, 3, 537-557. London: Reed Elsevier. Grechka, V., Bakulin, A. and Tsvankin, I. (2001). "Seismic Characterization of Vertical Fractures Described as General Linear-slip Interfaces." Abstract 71 st Annual International Meeting, 57-60 . Tulsa: Society of Exploration Geophysicists. Hall, D. H. (1976). "Crustal studies: Their Implications for the Explorationist." Exploration Geophysics, 07, 53-59. Sydney: Australian Society of Exploration Geophysicists. Hedlund, C. A., Couzens-Schultz, B. A., Maler, M. O., Garmezy, L. and Naruk, S. J. (1999). "The State-of-the-art in Interpretation of Complex Structure from 2D and 3D Seismic Data." Abstracts and Programs, Geological Society of America Meeting, 31, 7, 127. Boulder: Geological Society of America. Hesthammer, J. (1999). "Improving Seismic Data for Detailed Structural Interpretation." The Leading Edge, 18, 02, 226-247 . Tulsa: Society of Exploration Geophysicists. Hesthammer, J., (1998). "Evaluation of the Timedip, Correlation and Coherence Maps for Structural Interpretation of Seismic Data." First Break, 16, 5, 151-167. Houten: European Association of Geoscientists & Engineers. Hesthammer, J. and Fossen, H. (1997). "The Influence of Seismic Noise in Structural Interpretation of Seismic Attribute Maps." First Break, 15, 6, 209-219. Houten: European Association of Geoscientists & Engineers. Hockers, C. and Fehmers, G. (2002). "Fast Structural Interpretation with Structure-oriented Filtering." The Leading Edge, 21, 03, 238-243 . Tulsa: Society of Exploration Geophysicists. Huang, X., Sinha, B., Burns, D. and Toksoz, M. (1999). "Formation Stress Estimation Using Standard Acoustic Logging." Abstract 69 rd Annual International Meeting, 53-56 . Tulsa: Society of Exploration Geophysicists. Huber, W. (1988). "Thrust Fault, Gulf of Mexico (Ewing Bank area)." Bulletin Houston Geological Society, 30, 7, 17-18. Houston: Houston Geological Society.
Iran, B. (2003). "Structural Styles in the Zagros Simple Folded Zone." Journal of the Geological Society of London, 160, 3, 401-412. London: Geological Society of London. Ivanova, N., Verba, M., Sakoulinna, T. and Matveev, J. (2000). "Rift Structures of the Barents Plate from Regional Geophysical Investigation." Abstract 70 th Annual International Meeting, 709-712 . Tulsa: Society of Exploration Geophysicists. Jacobs, J. A. C., Jardin, A., Delprat-Jannaud, F., Versteeg, R. and Lailly, P. (1991). "Prestack Structural Interpretation of Seismic Data." Expand Abstract 61 st Annual International Meeting, 188-191 . Tulsa: Society of Exploration Geophysicists. Jackson, M. P. A., Schultz-Ela, D. D., Hudec, M. R., Watson, I. A., and Porter, M. L. (1998). "Structure and Evolution of Upheaval Dome: A Pinched-Off Salt Diaper." G eological Society of America Bulletin, 110, 1547-1573. Boulder: Geological Society of America Jones, G. and Knipe, R. J. (1996). "Seismic Attribute Maps: Application to Structural Interpretation and Fault Seal." First Break, 14, 12. Houten: European Association of Geoscientists & Engineers. Jones, P. B. (1988). "Seismic Interpretation 10 - Balanced Cross-sections - An Aid to Structural Interpretation." The Leading Edge, 07, 08, 29-31 . Tulsa: Society of Exploration Geophysicists. Jones, R. M. (2003). "An Integrated, Quantitative Approach to Assessing Fault-seal Risk." AAPG Bulleting, 87, 3, 507-524. Tulsa: American Association of Petroleum Geologists. Kattenhorn, S. A. and Pollard, D. (2001). " Integrating 3-D Seismic Data, Field Analogs, and Mechanical Models in the Analysis of Segmented Normal Faults in the Wytch Farm Oil Field, Southern England, United Kingdom." A.A.P.G. Bulletin, 85, 1183-1210. Tulsa: American Association of Petroleum Geologists. Kligfield, R., Rowan, M. G. and Ratliff, R. (1989). "Applications of Computeraided Section Construction, Restoration, and Balancing Models to the Petroleum Industry." Abstract 59 th Annual International Meeting, 589 . Tulsa: Society of Exploration Geophysicists. Koledoye, A.B., Aydin, A. and May, E. (2000). "Three-dimensional Visualization of Normal Fault Segmentation and its Implication for Fault Growth." The Leading Edge, 19, 07, 692-701. Tulsa: Society of Exploration Geophysicists. Konischev, V. (1998). "Rift Halokinesis." Abstract 60 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Konyukhov, A. I., Sokolov, B. A. and Yandarbiev, N. S. (1998). "Diapirizm, Mud Volcanizm and Folding in the Sedimentary Basins of Northern Peritethys." Abstract 60 th Annual Meeting. Houten: European Association of Geoscientists & Engineers.
Le Turdu, C., Keskes, N., Jeanjean, F., Kristvik, I., Ringenbach, J. and Straw, P. (1999). "New Tools for 3D Fault Detection by SISMAGE (super TM); Implications to Interpretation of Complex Faulted Zones of the North Sea." A.A.P.G. Bulletin, 83, 8, 1324. Tulsa: American Association of Petroleum Geologists. Looff, K. M. (2001). "Recent Salt Related Uplift and Subsidence at Sour Lake Salt Dome, Hardin County, Texas." Transactions, 51, 187-194. New Orleans: Gulf Coast Association of Geological Societies. McCabe, P. J. (1984). "Influence of Strike-slip Movement on Terrestrial Sedimentation in Upper Carboniferous of Nova Scotia and New Brunswick." Abstract 54 rd Annual International Meeting, 57-60 . Tulsa: Society of Exploration Geophysicists. Mehillka, L., Dhima, S. and Basha, J. (1999). "A New Option on Structural Styles and Independent Petroleum Systems in Outer Albanides." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers. Mitra, S . (1990). "Fault-Propagation Folding: Geometry, Kinematic Evolution and Hydrocarbon Traps." A.A.P.G. Bulletin, 74, 921-945. Tulsa: American Association of Petroleum Geologists. Mitra, S . (1988). "Effects of Deformation Mechanisms on Reservoir Potential in the Central Appalachians." A.A.P.G. Bulletin, 72, 536-554. Tulsa: American Association of Petroleum Geologists. Mitra, S . (1988). "Three-Dimensional Geometry and Kinematic Evolution of the Pine Mountains Thrust System, Southern Appalachians." Geological Society of America Bulletin, 100, 72-95. Boulder: Geological Society of America. Nilsen, T. H. and Sylvester, A. G. (1999). "Strike-slip Basins: Part 1." The Leading Edge, 18, 10, 1146-1152 . Tulsa: Society of Exploration Geophysicists. Nilsen, T. H. and Sylvester, A. G. (1999). "Strike-slip Basins: Part 2." The Leading Edge, 18, 11, 1258-1267 . Tulsa: Society of Exploration Geophysicists. Moretti, I., Delhomme, J. P., Cornet, F., Bernard, P., Schmidt-Hattenberger, C. and Borm, G. (2002). "The Corinth Rift Laboratory: monitoring of active faults." First Break, 20, 20, 91-97. Tulsa: Society of Exploration Geophysicists. Murtuzaev, I. and Bayramov, A. (1998). "Strike-Slip Faults of the South Caspian Basin - Azerbaijan and Adjacent Regions." Abstract 60 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Ott, V. D., and Eaves, G. P. (1998). "3-D seismic Examples of Structures from the Permian Basin." Abstract American Association of Petroleum Geologists 1998 Annual Meeting. Tulsa: American Association of Petroleum Geologists. Park, J. O., Tsuru, T., Kaneda, Y., Kinoshita, H., Kodaira, S., Kono, Y. and Takahashi, N. (1999). "New Seismic Reflection Images of the Nankai Subduction Zone Off Southwestern Japan from JAMSTEC Cruise Seamount Subduction and
Seismic Thrust Fault." Expand Abstract 69 nd Annual International Meeting, 10091012 . Tulsa: Society of Exploration Geophysicists. Parkin, E. L., Lockman, D. F. and Schwalbach, J. R. (1998). "Fault Detection and Analysis Using Borehole Image Logs at Pescado Field, Santa Ynez Unit, Offshore California." Abstract A.A.P.G. Pacific Section Meeting, 82, 5, 885. Tulsa: American Association of Petroleum Geologists. Pawlowski, R. (1999). "Megaregional Rift-drift Structural Controls on Hydrocarbon Accumulations Offshore West Africa." The Leading Edge, 18, 5, 600-603. Tulsa: Society of Exploration Geophysicists. Raditchev, R., Dimovski, S., Stavrev, P. and Gerovska, D. (1999). "Various Density Models of Rift Zones." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers. Ratliff, R., Rowan, M., Forester, R. and White, G. (1989). "Section Restoration and Balancing as an Aid yo Seismic Interpretation: GSI Line TWTS-83-13 (Regional Well-Tie Line, Offshore Galveston, Texas)." Abstract 59 th Annual International Meeting, 435 . Tulsa: Society of Exploration Geophysicists. Roberts, M.J. and Reilly, S.G. (2002). "Frontier Exploration on the Abyssal Plain of the Deepwater Gulf of Mexico - Syn-Rift through Passive Stage Mesozoic Objectives Identified." Abstract 64 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Rodnikova, R. D. (1998). "HC Resources of the Sea of Okhotsk Rift Basins in the Light of Fuel Demand in NE Asia." Abstract 60 th Annual Meeting. Houten: European Association of Geoscientists & Engineers. Rouby, D. and Suppe, J. (1996). "Restoration in 3D of Folded and Faulted Surfaces." Abstract Geological Society of America, 28th Annual Meeting, 28, 7, 241. Boulder: Geological Society of America (GSA) : Boulder, CO, United States Sabat, F., Muños, J. A., Poblet, J., Roca, E. and Verges J. (1997). "Bed-by-bed Fold Growth by Kink-band Migration: Sant Llorenç de Morunys, Eastern Pyrenees." Journal of Structural Geology, 19, 443-461. London: Reed Elsevier. Sayers, C. M. (2002). "Stress-dependent Elastic Anisotropy of Sandstones." Geophysical Prospecting, 50, 85-95 . Houten: European Association of Geoscientists & Engineers. Schneider, F., Bouteca, M. and Vasseur, G. (1994). "Validity of the Porosity/effective-stress Concept in Sedimentary Basin Modeling." First Break, 16, 2, 413-419. Houten: European Association of Geoscientists & Engineers. Scheidhauer, M., Beres, M., Dupuy, D. and Marillier, F. (2001). "A HighResolution 2D/3D Seismic Study of a Thrust Fault Zone in Lake Geneva,
Switzerland." Abstract 61 rd Annual Meeting. Houten: European Association of Geoscientists & Engineers. Schultz-Ela, D. D. (2003). "Origin of Drag Folds Bordering Salt Diapers." AAPG Bulletin, 87, 5, 757-780. Tulsa: American Asotiation Petroleum Geologists. Tang, X., Cheng, A. and Cheng, N. (1999). "Formation Stress Determination from Borehole Acoustic Logging: A Theoretical Foundation." Abstract 69 th Annual International Meeting, 57-60 . Tulsa: Society of Exploration Geophysicists. Toomey, A. (2003). "Seismic Monitoring of Stress on Fractures/Faults." Abstract 73 rd Annual International Meeting, 2263-2266 . Tulsa: Society of Exploration Geophysicists. Velaj, T. (1999)."Structural Styles and Exploration Opportunities from Thrustbelt to Platform in Western Albania." Abstract 61 st Annual Meeting. Houten: European Association of Geoscientists & Engineers. Viola, G, Odonne, F. and Mancktelow N. S. (2004). " Analogue Modelling of Reverse Fault Reactivation in Strike-slip and Transpressive Regimes: Application to the Giudicarie Fault System, Italian Eastern Alps." Journal of Structural Geology, 26, 3, 401-418. London: Reed Elsevier. Wickham, J. and Moeckel, G. (2002). " Sequential Restoration and Unstraining of Structural Cross Sections; Applications to Extensional Terranes; Discussion." AAPG Bulletin, 86, 1, 183-184. Tulsa: American Association of Petroleum Geologists. Withjack, M. O., Meisling, K. E. and Reinke, W. J. (1987). "Seismic Expression of Structural Styles." Abstract 57 th Annual International Meeting, 57-60 . Tulsa: Society of Exploration Geophysicists. Woodcock, N. H. and Rickards B. (2003). " Transpressive Duplex and Flower Structure: Dent Fault System, NW England." Journal of Structural Geology, 25, 12, 1981-1992. London: Reed Elsevier. Wride, V. C. (1995). "Structural Features and Structural Styles from the Five Countries Area of the North Sea Central Graben." First Break, 13, 10, 395-407. Houten: European Association of Geoscientists & Engineers. Zalan, P. V. (1987). "Identification of Strike-slip Faults in Seismic Sections." Abstract 57 th Annual Meeting. Houten: European Association of Geoscientists & Engineers.
Additional Reading
Allmendinger, R. W., T. A. Hauge, E. C. Hauser, C. J. Potter, S. L. Kemperer, K. D. Nelson, P. Knuepfer and J. Oliver. 1987. Overview of the COCORP 40¡N transect, western U.S.: The fabric of an orogenic belt. GSA Bull. 98: 308. Badley, N. E., T. Egeberg and C;. Nipen. 1984. Development of rift basins illustrated by the structural evolution of the Oseberg feature, Block 30/6, offshore Norway. Jour. Geol. Soc. London 141: 639. Brown, A. (1999). Interpretation of Three-dimensional Seismic Data. Tulsa: Society of Exploration Geophysicists. Chapman, T. J. and G. D. Williams. 1984. Displacement-distance methods in the analysis of fold-thrust structures and linked-fault systems. Jour. Geol. Soc. London 141:121. Cliffs, E. (1985). Principles of Structural Geology. New York: Prentice. Ellis, P. G. and K. R. McClay. 1988. Listric extensional fault systems-Results of analogue model experiments. Basin Res. 1(1): 55. Fangin, S. W. (1991). Seismic modeling of Geologic Structures. Tulsa: Society of Exploration Geophysicists. Fox, J. 1987. Seismic interpretation in salt-controlled basins. The Leading Edge (SEG) 6 (3): 10. Fulton, T. K. 1985. Some interesting seismic noise. 1985 Offshore Tech. Conf paper no. 4929, Houston, TX. Gibbs, A. D. 1984. Structural evolution of extensional basin margins. Jour. Geol. Soc. London 141: 609. Jarvis, G. T. 1984. An extensional model of graben subsidence-The first stage of basin evolution. Sed. Geol. 40:13. Johnson, J. D. 1992. Structural imaging in the real world. The Leading Edge (SEG)11 (1): 32. Jones, P. B. Balanced cross sections-An aid to structural interpretation. The Leading Edge (SEG) 7 (8): 29. King, G., and J. Nabelek. 1985. Role of fault bends in the initiation and termination of earthquake rupture. Science 228: 984. Mitra, S. and Fisher, G.W (1992). Structural Geology of Fold and Thrust Belts, Baltimore: Johns Hopkins University Press. Proffett, J. N., Jr. 1977. Cenozoic geology of the Yerington District, Nevada, and implications for the nature and origin of Basin-and-Range faulting. GSA Bull. 88: 247.
Roberts, A. N., G. Yielding and B. Freeman, eds. 1991. The Geometry of Normal Faults. Special Pub. no. 56: Bath, England: Geol. Soc. London Pub. House. Schlische, R. W. 1991. Half-graben basin-filling models: New constraints on continental extensional basin development. Basin Res. 3 (3): 123. Schumaker, R. C. 1981. Anatomy of a thrust belt as viewed by a petroleum geologist. Appalachian Petrol. Geol. Symposium, West Virginia Univ. Sclater, J. G. and M. D. Shoney. 1988. Mid-Jurassic through mid-Cretaceous extension in the Central Graben of the North Sea-Part 2: Estimates from faulting observed on a seismic line. Basin Res. 1(4): 201. Sclater, J. G. and B. Celerier. 1988. Errors in extension measurements from planar faults observed on seismic reflection lines. Basin Res. 1(4): 217. Tearpock, D. J. and Bischke, R. (1991). Applied Subsurface Gological Mapping. New York: Prentice Hall. Tearpock, D. J. and Bischke, R. (2002). Applied Subsurface Gological Mapping with Structural Methods. New York: Prentice Hall. Tucker, P. M. (1973). Pitfalls in Seismic Interpretation. Tulsa: Society of Exploration Geophysicists. Watts, A. B. 1982. Tectonic subsidence, flexure and global changes of sea level. Nature 297: 469. Woodward, N. B. and Boyer, S. E. (1985). An Outline of Balanced CrossSections. Knoxville: University of Tennessee Department of Geology.
