Baisic Geology Ieoc All

General Introduction

The area of geologic study known as tectonics, or sometimes geotectonics, is best thought of as structural geology in its broadest sense. It examines regional structural features-their interrelationships, evolution, and effects on sedimentation.

The concept of "structural style" is based on comparative tectonics. Its greatest utility lies in identifying certain basic patterns of deformation 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 rock, 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 geologists 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.

Classification of Styles

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: (1) compressional; (2) extensional; (3) strike slip (wrench); and (4) intracratonal (vertical). For the most part, these last two types of province are thought to involve basement in nearly every case.

Table 1 (below) lists the various structural styles currently used in exploration work.

Structural Style Dominant Deformational Force

Transportation Mode

Extensional

Extensional fault blocks Extension High to low-angle divergent

dip slip of blocks and slab

Detached normal fault assemblages ("growth faults" and others)

Extension

Subhorizonal to high-angle divergent dip slip of sedimentary cover in sheets, wedges, and lobes

Salt structures Density contrast Differential loading

Vertical and horizontal flow of mobile evaporites with arching and /or piercement

of sedimentary cover

Shale structures Density contrast Differential loading

Dominantly vertical flow of mobile shales with arching and/or piercement of sedimentay cover

Gravity structures Slope instability Differential loading

Downslope gliding on decollement

Compressional

Compressive fault blocks and basement thrusts

Compression High to low-angle convergent dip slip of blocks, slabs, and sheets

Decollement thrust- fold assemblages

Compression Subhorizontal to high-angle convergent dip slip of sedimentary cover in sheets and slabs

Decollement thrust- fold assemblages

Compresssion Subhorizontal to high-angle convergent dip slip of sedimentary cover in sheets and slabs

Strike Slip

Wrench fault Shear couple Strike slip of subregional to regional plates

Vertical

Basement warps: arches, domes, sags

Multiple deep-seated processes(thermal events, flowage, isostasy, etc)

Subvertical uplift and subsidence of solitary undulations

Structural Styles Primary Secondary

Extensional

Extensional fault blocks

Divergent boundaries:

1. Completed rifts

2. Aborted rifts : aulacogens Intraplate rifts

Convergent boundaries 1. Trench outer slope 2. Arc massif 3. Stable flank of

foreland and fore-arc basins

4. Back-arc marginal seas (with spreading)

Transform boundaries: 1. With component of

divergence

2. Stable flank of wrench basins


Passive boundaries(deltas)

Salt structures Divergent boundaries: 1. Completed

rifts and their passive margin sags

2. Aborted rifts; aulacogens

Regions of intense deformation containing mobile evaporite sequence

Shale structures Passive boundaries(deltas)

Regions of intense deformation containing mobile shale sequence

Gravity structures Passive boundaries Convergent boundaries:

1. Trench outer slope 2. Fore-arc basins

3. Back-arc basins

Compressional

Compressive fault blocks and thrusts basement

Convergent boundaries:

1. Foreland basins

2. Orogenic belt cores

3. Trench inner slopes

Transform boundaries (with component of convergence)

Decollement thrust-fold assemblages

Convergent boundaries:

1. Mobile flank (orogenic belt) of forelands

2. Trench inner slopes and

Transform boundaries (with component of convergence)

outer highs

Strike Slip

Wrench fault Transform boundaries Convergent boundaries:

Foreland basins

2. Orogenic belts Arc massif

Divergent boundaries:

1. Offset spreading centers

Vertical

Basement warps arches, domes, sags

Plate interiors Divergent, convergent, and transform boundaries

Passive boundaries

Table 1: Structural styles and their various plate-tectonic habitats

In our discussion, we will follow the basic organization in volumes II and III of A.W. Bally's excellent and extensive Seismic Expression of Structural Styles: A Picture and Work Atlas (1983). This series is itself a major contribution to contemporary comparative tectonics and should be consulted for more detailed and varied examples of each style.

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, it is the basement structure that 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, protooceanic (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 arc-continent and continent-continent collision zones;

3. transform, which includes ridge-ridge transform faults and wrench faults.

Many of the new ideas about the relationship between orogenesis and plate interactions focus on the events that occur along convergent boundaries. These boundaries have been divided into two main types ( Figure 1 , A- and B-type subduction. Two types of A- subduction are shown.).

Figure 1

"B," or Benioff-type, subduction is said to characterize regions where oceanic lithosphere dips under an island arc or continental margin. These boundaries are Cenozoic in age and have Benioff zones of shallow-intermediate-deep focus earthquakes. They most often "face" outward, i.e., oceanward, from the arc or continental margin beneath which lithosphere is subducted.

"A," or Alpine-type, subduction has been invoked to explain many of the world's foreland fold and thrust belts (e.g., the Canadian Rocky Mountains and Western Overthrust Belt, the Himalayas, the Zagros Orogenic Belt, the Alps, the Appalachians), which are characterized by continental crust dipping beneath a regional decollement fold-thrust belt, usually toward a high-grade metamorphic core. A-type subduction zones generally face continentward; i.e., the direction of thrusting

and overturning of folds within them is mainly toward the continental interior. They are primarily of Mid Paleozoic to Early Cenozoic age. Most examples have been explained as the result of continent-continent or continent- island arc collision, and thus as successors to B-type subduction. Several major provinces of this type, however-most notably the overthrust belts of western North America-have not been explained, and remain highly enigmatic in terms of their origin.

Because there are many more examples of B-type subduction than A-type, and since B-type is "in progress" everywhere, a considerable amount of specific nomenclature has evolved from recent analyses of its features and processes. The more important terms are shown in Figure 2 and Figure 3 ,

Figure 2

which portray the details of convergent margins (Figure 2: Regional setting of an active B-subduction zone, including a backarc (rifted) basin.

Figure 3

Figure 3: Enlargement of forearc region in Figure 2, showing common basin and trench nomenclature.). In general, because it appears that many of the world's major mountain systems have resulted from collisional events-i.e., they have evolved from B-type to A-type subduction-most researchers feel that an understanding of B subduction will help lead to more accurate and useful explanations of A subduction.

In our introduction to structural styles, we are by no means limited to discussing regions of mountain building. As we will see, five of the nine structural styles categorized thus far occur where diastrophic influence is comparatively slight, even absent. For most styles, we will examine both "evolving," and older, "completed" tectonic provinces. In addition, field examples will be given. In each case, it should be kept in mind that structure alone does not define hydrocarbon potential, nor does the existence of good reservoir rock. Certain regions, such as the Makran fold belt that we will discuss presently, have all the apparent prerequisites for accumulation, but may be characterized by low heat flow. Thus, a particular structural style must often be viewed in a larger tectonic framework in order to best evaluate its hydrocarbon potential.

Extension Styles – Basement Involved

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 as well as numerous, linear intracratonal grabens that occur on nearly every continent.

For these cases, block faulting is directly related to the process of rifting (i.e., continental breakup). In the 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.

Extensional provinces that involve basement in their deformation can be divided into two basic types:

Actual rift grabens and upwarps (whether active or failed).

Passive continental margins, which represent one side of a successful rift that has been buried by contemporaneous and subsequent nonmarine and marine sedimentation

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 exist.

Perhaps the most important characteristic that must be kept in mind with regard to 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, cuspate, or can alternate between these. 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.

Block Faulting in Rifts

Figure 1 shows three basic models that have been proposed to explain the geometry of rift-related faulting.

Figure 1

Much controversy continues to surround this subject. 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. According to Lowell et al. (1975), 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 this region are interpreted to show listric geometry at depth in cross section ( Figure 2 , Evolution of Red Sea graben.

Figure 2

Steepness of faulting is interpreted to decrease and listric geometry to predominate at depth) and a curving trace in plan view ( Figure 3 , Gulf of Suez graben: Part a is a block diagram of east central portion of the graben showing normal fault patterns.

Figure 3

Part b is a rose diagram fault plot for the same region.). Seismic profiles, however, have been interpreted to show more or less straight, relatively high-angle normal faults ( Figure 4 , Uninterpreted and interpreted seismic profile in western portion of the southern Red Sea.

Figure 4

Note tilting of lower sediments in each graben.), at least in several areas. Since reflection profiles generally exaggerated the curved geometry of faults at depth (due to velocity increase), the linearity of the fault planes in Figure 4 appears all the more distinct. The question of detailed fault geometry, therefore, remains partly unanswered in this region.

Fault-bounded blocks are often 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 hangingwall 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 recently been identified as accounting for about 10% of the total offset along basement block faults (Jackson and McKenzie 1983). Such uplift can lead to local thickening of sediments over the hangingwall block, thinning over the footwall block, and to local bathymetric highs that can localize reef growth or other shallow marine sedimentation.

Figure 2 shows the inferred structural evolution of the southern Red Sea rift as interpreted by Lowell et al. (1975). We might note that all stages-but particularly those that mark the beginning of tectonism-involve significant vertical uplift. The broad arching and thinning of continental lithosphere is accomplished by extensional adjustment in its upper, brittle portion (i.e., the crust) and is accompanied by high thermal gradients. This results in a central graben with flanking plateaus that remain elevated above surrounding terrane. These uplifts, therefore, deflect major drainage away from the embryonic ocean basin but contribute their own coarse detritus in the form of alluvial fans.

The general lack of clastic influx encourages the accumulation of a thick evaporite sequence. This can have two consequences of major importance to hydrocarbon entrapment: such evaporites can provide an excellent seal for underlying porous intervals and can, after subsequent burial, be mobilized into salt structures that create a host of trapping possibilities.

Figure 5

Figure 5

and Figure 6 show the distribution and nature of the major hydrocarbon

accumulations in the Gulf of Suez portion of the Red Sea.

Figure 6

Note the number of stratigraphic traps, including those associated with the unconformity separating the Cretaceous and Tertiary sections. This type of major unconformity between pregraben (generally Pre-Miocene) and graben fill deposits is typical of rift provinces and has considerable importance to exploration. Its overall significance is best discussed in terms of passive continental margins.

A look at the Piper oil field in the central North Sea graben ( Figure 7 ) reveals several other major features consistently seen in this setting.

Figure 7

Figure 8 (Productive fault blocks ),

Figure 8

Figure 9 (Structure contour map ),

Figure 9

Figure 10 (Cross section through Piper field),

Figure 10

Figure 11 (Interpreted seismic profile through Piper field),

Figure 11

and Figure 12 (Sedimentation model )are taken from Maher (1980) and show the relationship of production to block fault structure in the Piper field.

Figure 12

Note 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.

Rifting in this region began during the Early Mesozoic, as part of the crustal thinning and magmatic intrusion that led to the opening of the North Atlantic. By the end of Cretaceous time, however, the North Sea arm of the mid-Atlantic rift system had largely failed. The graben, therefore, has remained an intracratonic basin that has been below sea level since Mid-Cretaceous time.

Cross section and seismic data ( Figure 11 ) reveal how strata are draped over basement blocks. As shown in Figure 12 , these blocks exercised a dominant control over depositional environments. In particular, sedimentation of the high-energy, 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.

These bear detailed comparison to the structural features and patterns of this discussion. At the same time, their specific tectonic and depositional histories are quite different. For example, where the East African rift system has had sporadic activity over much of the Tertiary period, the basin and range has undergone almost 250 km of continuous extension since the Late Miocene-Early Pliocene Age.

In general, rifts have excellent hydrocarbon potential in comparison to other structural styles. Their evolution generates an abundance of source, migration, and trapping possibilities. These last relate not only to faults and the drape closures over them, but to salt sealing. Equally important is the existence of high heat flow, which is regional and continuous for millions of years. Fault intersections can result in trap-door blocks with overlying closure-an ideal structural trap. Drag folding along faults may create others. Unconformities created during the starved basin stages can also act as seals. The highest portions of rotated fault blocks can localize the growth of reefs or the deposition of shallow-water and porous sands, as at Piper field. Sediments deposited during these stages often include organic-rich lacustrine and swamp material. These form excellent source rocks.

Basement block faulting also occurs in settings of 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.

Extensional Structural Styles in Passive Margins

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. Passive margins are post-Triassic in age, having developed as a result of the breakup of the supercontinent Pangaea. They are characterized by seaward-prograding wedges of mostly shallow marine sediments that can reach 14 km in thickness (Bally and Oldow 1983). These deposits reflect in their overall character the basic stages of rifting and continental separation.

The ideal stratigraphic section (see Figure 13 , Idealized evolution of Atlantic continental margin from Triassic to the present, showing progressive burial of rifted basement structures.

Figure 13

The features shown are considered typical of most passive margins, Figure 14 ,

Figure 14

and Figure 15 Block diagrams illustrating the general succession of depositional environments that develop along passive continental margins characterized by mainly clastic deposition.

Figure 15

) begins with coarse, syntectonic, nonmarine clastics (alluvial fan, fan delta, braided stream), possibly interbedded with fine-grained, organic lacustrine shales. These sediments are commonly intruded by, and interbedded with, volcanic material, mostly of mafic composition. This sequence accumulates in half-grabens between major fault blocks, as well as within the central rift. It thus becomes broken up, tilted, and rotated as faulting continues.

Evaporites mark the earliest stages of subsequent marine incursion, as the rift expands to intersect an ocean basin. These may also occur at the base of the rifting section, if marine incursion occurs at the initiation of breakup. Thicknesses of salt may therefore be confined to halfgrabens, or may occur as more extensive sheets overlapping these local basins. As plate separation proceeds, and a new ocean is created, continental margin sedimentation begins and marine displacement results. A thick wedge of clastic-carbonate deposits then begins to prograde over the older, largely nonmarine sediments. In most cases, a regional unconformity separates these two sequences.

Basically, then, the evolution of passive margins can be divided into an early, rift-related 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).

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 the structures developed within them. The Cretaceous-Tertiary boundary (between units TP&E and K1 .4, K2) is marked by a sloping unconformity. This is interpreted to have been the result of submarine erosion, possibly related to changes in spreading rates and the relative height of sea level (see Vail et al. 1977).

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.

Figure 16

Figure 16

and Figure 17 show a map view and cross section (based on seismic profiles and data from several test wells) of the Baltimore Canyon area on the Atlantic margin of North

America.

Figure 17

Here, we see on the interpreted cross section how sediment loading has presumably depressed the transitional (i.e., between continental and oceanic) crust and accentuated the continental shelf edge into a hinge zone. This is typical for most passive margins. In addition, we might note the complex pattern of relatively long, narrow basins that exists both onshore and offshore.

Figure 18 shows a seismic section through this same general area of the Baltimore Canyon trough.

Figure 18

Here, note how deep the normal (growth?) faults extend, and also how the reefoid mass has been tilted strongly landward. This structural framework has been interpreted to be the result of plastic flow in a limy shale interval beneath the thick carbonate section. These shales were presumably rendered highly ductile by increasing overburden. With time, they were forced to flow basinward, thus removing material from behind the reefoid mass. This withdrawal of material rotated the overlying strata down, while simultaneously pushing the reefoid mass up, creating the large anticlinal structure and faulting seen. The resulting structural setting appears to offer several excellent potential structural traps.

Seismic reflection data also appear to indicate that many of the local basins shown in Figure 16 are actually half-grabens, separated by rotated fault blocks and joined transversely by transform faults (Bally and Oldow 1983, Harding 1984). Figure 19

Figure 19

and Figure 20 offer a close-up view of one such shallow basement half-graben filled by nonmarine deposits and covered by shelf sedimentation.

Figure 20

Such basins are presumably common features on nearly all passive margins. ( Figure 19 , Example of multichannel seismic profiles in the Baltimore Canyon area showing local half-grabens. Note tilting and deformation of synrift sediments within each half-graben and their general noncontinuous (nonmarine?) character. Figure 20 : Interpretation of depositional fill in Newark-type rift basins. Lacustrine (possibly swamp-derived in part) shales are highly organic. The petrophysical character of alluvial fan and fluvial-deltaic deposits is highly variable. This same type of sedimentary sequence is presumed to fill basins such as those shown in Figure 19 .)

As indicated by structures that have been studied in the Newark rift system of the eastern United States, the detailed history of displacement along the faults bounding these half-grabens can be quite complex, involving substantial amounts of strike slip motion. Figure 3 supports the notion that, due to the local variability of fault trends, these basins are also likely to be localized and, therefore, difficult to map along strike.

The nonmarine basin fill of the half-grabens of the Newark system is interpreted as shown in Figure 20 . The vertical and lateral juxtaposition of organic-rich lacustrine shales with alluvial fan and fluvial plain deposits may hold significant hydrocarbon potential.

Possible hydrocarbon traps in basement-involved fault blocks associated with passive margins include those mentioned for rift provinces in general. 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.

Detached Normal Faulting

Normal faulting that does not involve basement occurs both as a regional structural style in its own right, and as secondary deformation to other major styles. Cases of the latter include areas of extension on the crests of large folds, on the overhanging leading edges of major thrust sheets, or above diapirs involving salt, shale, or igneous intrusions.

By far the greatest number of hydrocarbon traps of this type have been found along passive margins in association with growth faulting, so named because of its syndepositional nature. Other common names for growth faults include "down-to-basin,"' 'contemporaneous," and, less frequently, "regional" and "down-to-the-gulf" faults. They are best developed in regions of deltaic or continuous clastic shelf deposition, where sediment accumulates rapidly and remains semiconsolidated to depths of several kilometers ( Figure 1 , Diagrammatic cross section through northern portion of the Gulf of Mexico, showing basic condition of metastability (due to density inversion) that leads to the consistent development of detachment structures).

Figure 1

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 2 shows an offshore growth fault, as interpreted on a seismic profile (Excellent example of a typical growth fault structure in the Gulf of Mexico).

Figure 2

Some of its more important characteristic features are:

the major fault plane is listric in shape, dips into the basin, and becomes parallel to bedding with depth;

displacement along the plane increases with depth;

the thickness of individual units increases abruptly in the hangingwall, due to syndepositional offset;

both secondary synthetic and antithetic faults complicate the basic structure. Many of the former sole out along the main detachment plane;

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);

in map view, the fault plane is cuspate, and is usually concave toward the basin center; antithetic faults, however, often show the reverse geometry.

Figure 3 shows the most common hydrocarbon traps that result from growth faulting.

Figure 3

As with rift-related structures, cross sections appear relatively simple, but in map view, faults intersect, die out, splay, and generally vary along strike to create complex patterns ( Figure 4 ).

Figure 4

In Figure 5 , a dip log from an offshore Louisiana oil and gas field shows the distinct patterns of growth faulting at two levels. Note the upward-decreasing dip motifs of the two rollover zones.

Figure 5

Also apparent is a highly consistent azimuth trend to the northeast, except in the uppermost shales. This is most likely due to tilting associated with the faults.

The flattening of faults with depth has been variously interpreted. Geologists 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 it to increasing compaction: faults originally develop at steep angles that are flattened 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 6 and Figure 7

Figure 7

, faults flatten out into shale-dominant prodelta slope facies (Figure 6: Generalized line drawing based on seismic profiles in the Gulf of Mexico showing relationship between depositional facies and growth fault development.

Figure 6

Figure 7: Diagrams illustrating the proposed development of growth faulting as a result of local "sinking" of sand-rich facies into underling clay-rich sediments.).

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 8 , which shows a complex example of large-scale, growth-type faults developing above the deepest portion of a rifted margin (Generalized cross section though Baltimore Canyon trough, showing detached faulting above presumed flowage of limey shales, A and B).

Figure 8

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 (see also Figure 9 , Interpreted seismic profile through the Baltimore Canyon trough, showing many features associated with passive continental margins. Note lump of material that has moved downslope under the influence of gravity.).

Figure 9

Some researchers consider gravity to be the principal cause of growth faulting in areas such as the Gulf Coast or Niger deltas. With respect to gravity, the maximum principal stress is vertical and it is the shear component of this stress, as resolved along the regional depositional slope, that 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 10 and Figure 11 , meanwhile, show the fallacy of trying to strictly segregate structural styles in every situation.

Figure 10

(Figure 10: Migrated time seismic section showing structure of the Hibernia discovery area, approximately 250 miles due east of Newfoundland.

Figure 11

Note scale of growth-type faulting and how it cuts deeply into basement. Figure 11: Time-seismic structure map on lower Cretaceous marker horizon, showing basic structure of the Hibernia anticline.) The Hibernia discovery was made approximately 250 miles due east of Newfound-land in a narrow, Late Mesozoic graben (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 11 ), is growth faulting, with large-scale rollover (compare Figure 2 and Figure 10 ). The faults, in this 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 (growth faulting) also appears in the features of a basement-involved style (rift fault block).

Salt Structures

The upward movement of salt in the form of diapirs within sedimentary sequences has long been assumed to be caused by the lower density of salt relative to the surrounding and overlying rock (Nettleton 1934). This density difference renders deeply buried salt buoyant and mobile. Due to this general metastable condition, its diapiric rise-often called halokinesis ("salt movement")-may or may not be initiated by tectonic influence. On the other hand, diapirism is, itself, a deformational force, piercing, drag-folding, and faulting the sedimentary layers through which it penetrates.

1. In general, the original deposition of salt is very often associated with rifting, which, in its initial stages, creates relatively narrow grabens with restricted circulation. Thus, salt structures are common along passive margins and in rift zones. Most often, the thickness of deposited salt increases into the host basin or graben depocenter. There is a progressive change in both the style of salt mobilization and its effect on the overlying sedimentary cover, which has been described as follows (refer to Figure 1 , Principal types of salt structures; contours in arbitrary units)

Figure 1

2. A withdrawal of salt from the original depositional edge, with a consequent downbending of overlying sedimentary layers

3. A contemporaneous pillowing of salt that may result in isolated swells or low, continuous ridges

4. The formation of low domes and anticlines that can have sufficient structural relief to pierce overlying sedimentary layers and form local salt stocks

5. The formation of piercement salt walls, diapirs, and detached features

Two well-studied successions of actual salt structures, showing the variation from basin margin to center, are shown in Figure 2 (Salt structures of northwest Germany.

Figure 2

Salt is from the Permian Zechstein group, a synrift depositional sequence) and Figure 3 (General inferred relationship between evolution of a basin and that of salt structures for a prograding continental margin).

Figure 3

Note the basin-ward increase in the size of individual structures. Figure 3 relates the style of diapirism and resulting salt structures to the rate of sedimentation and, therefore, basin evolution. As shown, continued regression along a continental margin would presumably cause an evolution of diapirism from swell structures to spines.

The overall orientation of salt diapirs, however, does not always reveal a consistent relationship to basin margins (see Figure 4 , Distribution of salt structures in the northwestern Gulf of Mexico).

Figure 4

Where uninfluenced by tectonism, diapirs tend to cluster somewhat randomly, or in subparallel ridges whose alignment may be determined by overburden. In cases where mobilization results from tectonic influence (e.g., rejuvenated basement faulting, or thrusting or folding in collision zones), patterns should appear that roughly indicate deformational trends (Trusheim 1960). Due to its extremely low ductility, salt (and evaporites in general) can serve as a plane of decollement. In mobile belts, it has often served to localize deformation, forming the cores of large, flexural-slip anticlines whose limbs become at least partially detached along contacts with the salt intervals.

A hypothesis that refines the traditional model of diapirism has been proposed by Bishop (1978). In this view, salt is mobilized as a series of waves flowing both basinward and vertically upward before a front of prograding overburden. Such a scenario can be inferred from Figure 5 , which apparently shows the main thickness of salt pushed up in front of the great weight of Late Mesozoic and Tertiary deposits (Regional NW-SE cross section through Gulf of Mexico).

Figure 5

The traditional concept of salt diapirism has assumed the upward movement of large masses of salt through a thick overburden. Despite its broad acceptance, this concept suffers from what might be termed a space problem. Basically, it does not adequately predict what happens to the column of material displaced by the salt. The observed folding and fault displacement along the margins of a diapir do not appear to account for the total space occupied by many piercement diapirs.

In a discussion of this problem, Woodbury, Murray, and Osborne (1980) propose that salt diapirs do not "jam" their way upward like volcanic plugs, but rather remain at about the same vertical position relative to the sedimentary cover. As this cover thickens by normal deposition, and the relevant basin subsides, a diapir presumably builds downward by drawing salt from the mother bed. The primary motion, therefore, is of sediments sinking past the salt body. Differential subsidence is assumed to create tension and normal faulting above the diapir. Figure 6 shows a comparison between the concepts of upward-building and downward growth of diapirs (Sequence a: proposed for structures in northern Germany, shows active intrusion of salt through overlying sedimentary cover.

Figure 6

Sequence b: proposed for the Gulf Coast, the "diapir" remains at the same height relative to the earth's surface and grows downward during concurrent sediment loading basin subsidence.).

Overall, individual halokinetic structures are extremely variable. Their development, location, and relationship to structures in the basement, as well as the character of the overlying sedimentary cover, are determined by the specific architecture and history of the basin in which they were deposited.

Extension above salt structures causes anticlinal warping and normal faulting. The latter is sometimes termed keystone, for reasons that should be apparent in Figure 7 (Seismic section of salt diapir showing keystone normal faulting in overlying units).

Figure 7

In map view, the faults are most often radial in their pattern ( Figure 8 , Structure map showing complex radial fault system associated with a productive salt dome, offshore Louisiana.

Figure 8

Note greater well density in area of particularly complex faulting, where numerous separate blocks exist.). However, as with normal faults in other structural styles, local complexity can be great. Antithetical faulting, for example, is also common.

Figure 9 shows a series of salt structures developed in the Aquitaine basin of southwestern France, on the eastern passive margin of the Atlantic.

Figure 9

Sediment thickening and thinning due to diapirism, as well as anticlinal warping and the development of a major unconformity, are all shown as the consequences of salt mobilization. In addition, the loci of mobilization appear to be over earlier, rift-related basement faults.

The basic scheme of hydrocarbon traps associated with salt structures is shown in Figure 10 (1=anticline, 2=graben caused by extension, 3=porous cap rock, 4=flank sand pinchout, 5=overhang, 6=nonoverhanging wall of structure, 7=angular unconformity, 8 and 9=normal faulting along flank of structure).

Figure 10

In many cases, the largest traps are in reservoirs not pierced by the salt; for example, numbers 1, 2, 8, and 9 in the figure. This, however, is not a hard and fast rule.

Shale Structures

The conditions that encourage the flowage of shale within a thick wedge of clastic deposits are more specific than are those for salt. They 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 even more impermeable upper seal 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 "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" (see Figure 1 , Concept of depositional "spoon," composed of compactible, sand-rich half-spoon overlying noncompacible shale-rich sequence), 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 in-to 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 basin-ward. 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 pro-gradation continues, and, thus, the generation of other growth faults ( Figure 2 and Figure 3 , Evolution of shale mound through growth faulting).

Figure 2

They may also rise as diapirs into overlying units or they may spread laterally.

Figure 3

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 3000 m in height, with a trend that is roughly parallel to regional strike ( Figure 4 , part (a) Seismic expression of shale structures.Shale mounds in northern Gulf of Mexico.

Figure 4

Shale diapirs also seem to be related to gravity sliding along master growth faults, which become planes of detachment at depth ( Figure 4 , part (b) Relationship between growth faulting and shale flowage).

Shale structures are not usually of great significance to petroleum exploration (the Beaufort Sea represents an important exception). However, they are 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.

Gravity Sliding

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 decollement thrusting of the Canadian cordillera, the chaotic sedimentary sequences known as "melange" (Fr. "mixture") or "olistostromes," the disturbed layering of single depositional layers, and the generation of growth faulting.

Gravity structures can very broadly be divided into those that are presumably related directly to orogenesis and those that are generated as a result of normal deposition. De Sitter (1964) and De Jong and Scholten (1973) offer comprehensive summaries of the role that gravity has been thought to play in the tectonics of foreland belts. Much of this, however, has become obsolete because of recent advances in plate tectonic analysis and theory.

Figure 1 (Gravity tectonics as envisioned for the creation of structures observed in the Apennines of central Italy.

Figure 1

This generalized evolution shows the basic features previously thought to cause foreland fold and thrust belt structures.) is a simplified diagram that illustrates the basic assumptions of gravity tectonics as a mechanism for thrust generation. The principal reasons such a mechanism fails to explain how laterally extensive sheets of sedimentary cover are transported toward a cratonal interior are as follows:

1. Basement slopes are usually in the wrong direction.

2. Regions of "tectonic denudation" are not generally seen.

3. Predicted thrust development is chronologically reversed from what is actually observed in foreland belts.

4. Sliding off uplifted regions would inevitably create relatively local tectonic aprons incompatible with the vast, laterally continuous thrust belts that parallel plate boundaries for thousands of kilometers.

5. No such mechanism is observed in any of the world's presently active foreland belts.

As a result, gravity is now invoked to explain specific structures of lesser scale. 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, slower-moving rotational blocks that show growth fault-type structures at their head. These we will review in a moment. 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).

Certain authors have pointed to the utter disorder in melange terranes as explainable in terms of gravity sliding. The jumble of variously sized sandstone, carbonate, mafic, and ultramafic (i.e., oceanic crustal) blocks within a matrix of shales that show signs of intense shearing seems to indicate the type of plastic flow and general loss of cohesion that we should expect in a mass that has moved relatively rapidly downslope. Many melanges are also divided by thrusts that appear to separate individual "slide" masses, a phenomenon that is predicted by models such as that in Figure 1 .

More recent interpretations, however, consider melanges to be the result of deformation at the leading edge of overriding plates at subduction zones (Karig 1974; Seely and Dickenson 1977). As shown in Figure 2 (Creation of melange.

Figure 2

Shallow structure and process as inferred from seismic profiles and field work in the eSumatra region. Gravity sliding takes place down the trench slope break, into the trench ) this explanation proposes the tectonic mixing of lithified sediment riding the subducting crust, as well as splinters of that crust, into the fine-grained, semiconsolidated material deposited in the trench. Gravity sliding in the form of submarine slumps down into the rise of the trench slope break are also thought to occur, but the principal mechanism of melange creation is the "scraping off" and incorporation of subducting material from the downgoing plate.

The specific character of a melange, in this scheme, will depend largely on the nature and amount of this material, its initial structural features, the rate of subduction, and the dip of the subducting lithosphere. Given the variety in each of these parameters throughout the world, melanges might be expected to show greater similarity in their overall style of deformation than in their specific characteristics.

Many geologists now believe that the generation of certain foreland fold and thrust belts is a result of passive margin clastic-carbonate wedges becoming involved in accretionary prism-type diastrophism during the early stages of continent-continent collision. If this be the case, the direction of tectonic transport would be toward the subducting plate (see Figure 3 (b) The megasuture is due to continental collision and thus represents a successor from B-subduction).

Figure 3

In relation to normal deposition along passive, divergent margins, so-called thin-skinned 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 4 ,

Figure 4

Figure 5

Figure 5

and Figure 6 ( Figure 5: A proposed evolution for gravity structures interpreted to exist in Zone 4a, Figure 6 : Proposed evolution for gravity structures interpreted to exist in Zone 4b.

Figure 6

) These have been derived from detailed study of the Mexican Ridge area in the western Gulf of Mexico. ( Figure 7 , Tectonic map of southern Mexican ridges area. Note convex-outward pattern of structural trends.) shows the distribution of the relevant structures.

Figure 7

Note how these vary between zones 4a and 4b, and how the two proposed evolutionary schemes attempt to explain this change. Zone 4b, which is characterized by growth faulting, has a slightly steeper 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 "scarps" 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 8 (Rollover due to gravity gliding in Texas Gulf coast.

Figure 8

Trapping is due both to rollover and fault sealing.) The irregularity of the glide plane ( Figure 9 , Large -scale cross section showing regional structure that includes the features shown in Figure 8 ), which is partly due to shale diapirism, acts to localize the degree of bed rotation.

Figure 9

Movement along the basal detachment is assumed to be by continuous gravity creep. Note how 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 4 , Figure 5 and Figure 6 faulting should be considered potential hydrocarbon traps in deeper-water areas off passive margins.

Compressional Styles - Basement Involved

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 30º. 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 characterized by A-type subduction, and in the inner trench slope and outer high portions of B-type 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 ( Figure 1 , Principal basement uplifts and associated basins of the Rocky Mountain foreland region Cross section lines refer to Fig 3).

Figure 1

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 basin(s), 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 accumulation and subsequent 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 2 (Basement faulting in the Rocky Mountain foreland region) shows four styles of faulting that have been proposed to explain the geometries observed in the field and in seismic profiles.

Figure 2

To some degree, the differences between these reflect the fact that various investigators have concentrated their efforts in different parts of the 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 30º or less. ( Figure 3 , Two interpreted seismic profiles from Wyoming showing expression of rigid basement uplifts.

Figure 3

(a) Southwest Wind river fault; (b) Casper Arch thrust. Note how these "end member" styles of faulting correspond to (b) and (d) of Fig.2) Single faults apparently show both these geometries along their strike. At present, the proposed styles of faulting reflect two major schools of thought. These 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 2 parts (a) and (b), or horizontal "overthrusting" of parts (c) and (d). The seismic profiles given in Figure 3 illustrate how apparently irreconcilable these two schools are.

Many of the principal faults have also apparently suffered a degree of ancillary strike slip displacement. For these reasons, some geologists now think that at least several of the major blocks, such as the Wind River, have been uplifted in a type of scissorlike 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 (see, for example, 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:

a tilted fault block of basement and immediately overlying units.

an intermediate level where the principal fault becomes a zone of steep drag folding.

an uppermost level of gentle drape or monoclinal folding.

The structure of fault segments can be complex, involving normal faulting along the crest of the block and drape fold, tear faulting of the cover and basement, and more minor reverse faulting in the foot-wall of the main fault.

Some geologists have pointed to older zones of Precambrian shearing within the basement itself as a probable control on the location and possibly geometry of faults. Certainly, in any region where basement faulting is a dominant structural style, the geologist should 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.

Figure 4 (Regional east-west cross section through the Rocky Mountain foreland, from northeastern Wyoming to southeastern Idaho.

Figure 4

The deep crustal faulting shown for the Wind River Mountains is based on recent COCORP seismic reflection profiles. Note that faults are interpreted to reach to the crust - mantle boundary. Western end of section shows decollement thrusting of the western overthrust belt) shows basement faults as both relatively planar and as listric to the crust-mantle boundary. Compressive decoupling is, therefore, proposed to occur at or near the Moho. At present, this interpretation is highly speculative. It does, however, take account of the prevailing hypothesis of intracontinental underthrusting (a form of A-type subduction).

The most generally accepted idea about the genesis of these basement foreland uplifts is that the principal compressional stresses are in some way directly related to B-subduction, as appears to be the case in the South American cordillera ( Figure 5 , Basic tectonic setting and regional cross section through northwest Columbia, showing three principal tectonic divisions.

Figure 5

Note the extensive involvement of continental basement in faulting of the eastern cordillera). Here, an A-B subduction couple seems responsible for the opposing directions of tectonic transport.

Broadly speaking, the vertical rise of large, rigid tectonic blocks has generated a good deal of local variability, and thus a considerable diversity of structural traps

should be expected ( Figure 6 , General hydrocarbon trapping possibilities associated with basement block uplifts).

Figure 6

The Elk basin oil field in the Big horn basin of northern Wyoming is an example of one of the larger traps discovered thus far. To date, it has produced over 500 million bbl out of fault-controlled closures such as those shown in Figure 7 (Uninterpreted and interpreted seismic sections through South Elk basin producing area (northeast Big Horn basin, Wyoming), showing anticlinal folding over basement thrusts.

Figure 7

Note how deformation in this section appears to correspond with style (c) in Fig 2) and Figure 8 (Cross section of Elk Basin field approximately 10 miles north of the seismic profile shown in Fig 7 ).

Figure 8

In more recent years, some companies have attempted to penetrate several of the lower-dipping Precambrian thrusts in order to explore the sedimentary section beneath, certain formations of which are known to be productive in nearby domes and folds. To date, only a few of these expensive wells have been successful. Oil and gas leaks within the thrust zones, however, have been reported.

More generally, in addition to providing drape and fault-related closures, basement uplift can also influence petroleum generation and trapping in more subtle ways. For example, in the Hardeman basin of northeast Texas, high-angle splinter faults related to late Paleozoic uplift of the Red River-Matador Arch and the Wichita Mountains created local avenues for the invasion of dolomitizing solutions into the cores and flanks of Mississippian bioherms. As a result, these became excellent, though highly localized, reservoirs.

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 into high-angle reverse faults or thrusts. Such reactivation, since it reverses the sense of displacement. is said to create "inverted" basins. This can be crucial to understand 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.

Compressional Styles - Basement Not Involved

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 (shows basal decollement and telescoped nature of regional thrust deformation.

Figure 1

Note the interpreted involvement of matamorphism in thrusting toward the west, i.e., closest to the hot, mobile core of the orogen ) 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. The majority of hydrocarbon production exists in the foothills region, where faulting is especially complex. Figure 2 is a cross section through the Jumpingpound gas field which amply shows this complexity.

Figure 2

Note the degree of imbrication above the ramp anticline, where closure is best: this is the primary trap in this and many other fields of the foothills area.

Thrusted 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. We have already discussed some of the details of B-type subduction zones. Figure 3 displays the interpreted structure on a seismic section across the active subduction complex offshore of northern California.

Figure 3

This section shows the decollement faulting in the accretionary prism quite well.

Note the change from the chaotically deformed accretionary prism sediments into the more conformable forearc basin at the far right. The deformed complex is broken into thrust slices about 400-500 m (1300-1600 ft) 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 verge toward the subducting ("on-coming") plate and represent the succession of B-subduction by A-subduction. ( Figure 4 , a map showing distribution of the major collision-related fold and thrust belts and related foredeeps (regional foreland basins.

Figure 4

) Many of the world's foredeeps (Appalachian, Canadian Rocky Mountain, Arabian, Uralian) are highly productive of oil and gas. Cross section lines refer to Figure 5. ) A transition to basement-involved thrusting often occurs in these settings behind the thrust front. Figure 5

Figure 5

and Figure 6 (Generalized cross section through four of the world's major collision-generated mountain systems at comparable scales) show how this involvement is usually interpreted to increase into the mobilized,

Figure 6

metamorphic "core"' of an orogen, such as the Canadian cordillera. 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, then, 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" (see Figure 4 ). The cause for this type of variation along tectonic strike is not well understood, but often thought to be related to preexisting basement features.

Figure 5 and Figure 6 compare the overall features of several major collision-generated foreland regions. The Alps appear to have resulted from a series of collisional episodes between the Eurasian continent and various arclike continental fragments that once bordered it to the south. The Zagros orogen, as we have noted (see Figure 7 , Generalized map and cross section showing continental breakup along the Red Sea rift and collision in the Zagros region of southeastern Iran.

Figure 7

Numbers indicate total estimated separation (in km) between Africa and Arabia), is the continuing result of Eurasia colliding with the much smaller Arabian continent. Both the extensive Himalayan Mountain belt and the Appalachian-Ouchita-Marathon system are interpreted as the result of the impact between larger continental masses-in the first case, India and Eurasia, in the second, those Paleozoic continents (proto-North America, South America, Africa, and Europe) whose collision marked the early formation of Pangaea. In both the Alps and the Himalayas, great thicknesses of crystalline basement rocks are involved in the deformation and have exerted considerable control on resulting structures. Though decollement thrusting is evident in the Jura Mountains of western Switzerland and eastern France, it does not dominate the regional structural style of the Alps as it does the Appalachian, Zagros, or Canadian forelands.

At the same time, however, we can also think of the Alps and the Himalayas as two opposite end-members with respect to the general style of deformation The the sedimentary cover. The central and southern European Alps, with their spectacular development of thrust and fold nappe structures, represent the most highly contorted foreland region in the world. Broadly speaking, they show the effects of very rapid, collision-related diastrophism on a thick, only semicompacted sedimentary pile. Folding of a highly ductile nature is common; recumbent and overturned nappes are piled up on top of each other like rumpled carpets to form the higher ranges (see Figure 8 , Alpine nappe structures). Sediments appear to have

been squeezed up and out of forearc and backarc basins, as the various island arc fragments were progressively welded to the Eurasian continent.

Figure 8

The Himalayas present a very different case. Here, great thicknesses of well-indurated Paleozoic and Early Mesozoic sandstone and carbonate rocks were involved in the deformation. This generally created less contorted, more widely spaced, and far more massive structures.

As in the Canadian Rocky Mountains and the western overthrust belt of the United States, ramping is well developed in the Appalachian foreland. This has been interpreted as being the result of both ductility contrast related to stratigraphic variation and preexisting block faulting in the basement.

The well-exposed, dramatic structures of the Zagros foreland appear to resemble those of the Appalachians more than those of the Alps or Himalayas. Some of the larger folds stretch for as much as 160 km (100 miles) along strike before plunging beneath the surface. The existence of salt layers in the lithologic section has resulted in a high degree of complex, local decollement. In places, for example, slip has occurred along and within salt intervals. This has apparently created shallow anticlines that have subsequently been pierced by the tectonically mobilized salt.

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 hangingwall, 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 fore-land basins (sometimes called "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 out in front of the thrust belt, presumably created by crustal loading on a more massive scale.

Thus, deformation appears to begin in the deeper, thicker portions of the sedimentary wedge and progress upward and on-to the craton. In a broad sense, this has meant that hydrocarbons have had the best chance to accumulate and remain undisturbed near the youngest, leading portions of thrust belts and in the regional foreland basins out in front of them.

As mentioned, traps in thrust belts are mainly associated with asymmetric anticlines ( Figure 9 , Seismic profile through the eastern Po plain in northeastern Italy (approximately 50 km southwest of Venice), showing thrust structure of the Apennines.

Figure 9

This profile reveals the thrust belt at its widest point. Gas pools exist in Pliocene sandstones that wedge out against the rising structures. Oil is produced from complex structural traps in underlying Mesozoic carbonates) and, to a lesser extent, fault truncations. Closures are most often at less than about 3000 m depth (Harding and Lowell 1979). The actual size of individual traps, and their height of closure can vary a great deal, depending on the spacing of thrusts and the degree of asymmetry in folds.

Substantial-even giant-accumulations have been discovered in a number of the world's foreland regions. Perhaps the most impressive example of production from thrust-fold structures is offered by the oil fields in the Zagros Mountains. Here, reservoir quality is due to an extensive, interconnected fracture system generated by the folding. Production is from the Asmari Limestone, which, by itself, yields more than 75% of all petroleum currently being recovered from traps in foreland belts. In addition, several recent major discoveries along the Idaho-Wyoming thrust belt have encouraged continued exploration and have caused geologists to take another, more detailed look at the petroleum potential of other foreland regions, such as the Appalachian-Ouchita system.

With regard to active subduction zones, several general statements can be made concerning overall hydrocarbon potential. In the forearc, for example, both source and seal can be more than adequate, but the large amount of volcanogenic material

generally makes for rather poor reservoir quality. Not only is the sediment matrix often fine-grained, but both primary and secondary pore-plugging and swelling clays-most notably illite and montmorillonite-are abundant.

Moreover, low heat flow is characteristic of the forearc; this is even more the case in the forearc basins that develop over the subduction complex itself. Structural traps, however, abound and consist of anticlinal and thrust closures and relatively shallow drape folds above thrusts.

The problems mentioned for forearc regions appear to characterize the Makran subduction complex, which, because of its unique setting, may otherwise appear to offer relatively strong hydrocarbon potential ( Figure 10 ,

Figure 10

Interpreted structure of the Makran accretionary prism, Gulf of Oman,see Figure 4 for approximate location) This very large accretionary prism is basically the continuation of the Zagros collision zone into a B-type subduction zone that stretches nearly 900 km eastward from the southwestern coast of Iran to Pakistan. It shows the development of many thick, coherent thrust-fold structures whose amplitudes are unusually large and whose petroleum potential might therefore seem to be substantially greater than that in most other forearc arc systems. As much as 7 km of mostly late Tertiary abyssal plain sediments (most likely resulting from high erosion

rates that began with the India-Eurasia collisional event to the north and west) have been deformed into an imbricated stack.

Despite this structurally attractive setting, potential appears moderate at best due to relatively low geothermal gradients, which average 1 º per 30 m (Harms et al. 1983). Gas seeps and traces of heavy hydrocarbons, however, have been found. At the same time, wells drilled in coastal Makran have encountered very high pressures. Because of the young age of the sediments penetrated (Pliocene at total depth), their rapid accumulation (about 300 m per my), and their subsequent deformation, we can expect pore pressures to be quite high. As a whole, then, the combination of remote access, potential drilling problems, and low-to-moderate maturation potential makes most active forearc systems relatively high exploration risks at present.

Strike Slip Tectonics: 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. In all cases, they are assumed to involve basement. These fault systems can be pure strike slip or they may include components of compression (sometimes called ""transpression") or extension ("transtension") ( Figure 1 , Diagram illustrating the evolution of various structures associated with major wrench faults. Arrows labeled "C" and "E" indicate compressional and extensional components).

Figure 1

Major wrench faults are usually associated with transform plate margins. In our general discussion of faulting, we have looked at several major examples of these structures ( Figure 2 and Figure 3 , (Examples of major strike slip faults in various parts of the world.

Figure 2

Dots mark the site of active volcanoes.) which can occur as single shear planes or systems of parallel faults (Harding 1974,1983, 1985).

Figure 3

For petroleum exploration purposes, the geologist is most interested in the structures associated with wrench faults, for it is these that 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 strike slip 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. Figure 4 (Diagram illustrating the evolution of various structures associated with major wrench faults.

Figure 4

Arrows labeled "C" and "E" indicate compressional and extensional components) shows how they 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.

An example of a productive backarc wrench fault setting in Sumatra is given in Figure 5 (Map of Sumatra showing location of the Barisan wrench fault and associated folds and oil fields.

Figure 5

Large arrows indicate relative direction of magnitude of plate convergence in the Java Trench. Box at lower right indicates area of seismic line shown in Fig. 6.) and Figure 6 (Inverted basin structure - a case of superposed tectonism.

Figure 6

Note the differences in offset indicated by upper reflectors versus lower reflectors. Reactivation of basement faulting in compression is presumed to be responsible for the structural inversion observed. The Rambutan field produces from fold closures in the hangingwall of an inverted fault.) In this case, wrenching has created a series of large, secondary anticlines and inverted basins. Reservoirs are in Late Tertiary sands derived from a crystalline source. Notice that for several of the major faults shown in Figure 7 (Interpreted seismic profile across a wrench zone in the Andaman Sea, showing a negative flower-type structure.

Figure 7

Note that seismic data is relatively chaotic immediately east of the interpreted fault. Letters "A" and "T" refer to the blocks whose relative motion is away from or toward the viewer), the original sense of offset is preserved at depth, while the upper layers reveal the latest displacement.

Both folds and reverse faults are commonly associated with wrench faults characterized by transpression. The strike of these secondary features is commonly at low angles (not parallel) to that of the main fault plane. They also frequently occur in echelon patterns (as diagrammed in Figure 4 ). Extensional structures are primarily normal faults. As we have seen in Figure 5 and Figure 6 , transpression can also cause basin inversion and thus relatively shallow closures.

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. Figure 7 gives an interpreted example of a negative flower.

Figure 8 (Structure contour map of Los Angeles basin, showing location of Whittier and Norwalk faults.

Figure 8

These are compressional structures, related to the San Andreas fault system. As shown, oil fields in this area are associated with both fold and fault closures ) and Figure 9 (Cross section through Whittier oil field, showing complex structural relationships.

Figure 9

Whittier fault zone is highly mylonitized and acts as an updip seal. Note apparent positive flower structural style ) together show an example of the complex but productive structures associated with wrench faulting in the Los Angeles basin of southern California. We should note in the cross section that the Whittier fault, with its highly sheared mylonite zone, acts as an updip seal. The general appearance of the structure is a positive flower, with several generations of faulting evident. This type of complexity is common along wrench fault zones, and can create a variety of potential traps ( Figure 10 , Trapping possibilities associated with wrench faults). In addition to anticlinal and fault-trap closures, shearing on such a scale may be expected to generate fractures in surrounding competent lithologies (Harding 1985).

Figure 10

Despite the relatively wide use of flower structure terminology, questions remain as to its value. Often, seismic data in the crucial core areas of supposed flowers is inconclusive and open to 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.

Vertical Tectonics: Basement Warps

The deep interior of every major craton is characterized by broad regional arches and open circular basins that can contain more than 14 km of sedimentary fill. Figure 1 (Regional NW-SE cross section through Gulf of Mexico) shows the distribution of the world's principal intracratonal basins.

Figure 1

These structures have proved to be the sites where vast quantities of hydrocarbons have accumulated in both structural and stratigraphic traps. Such traps are often far greater in continuous extent than are those in more tectonically disturbed regions. To date, plate tectonic theory has been unable to adequately account for the existence and behavior of such major intracratonal warps. It would seem at this point that accurate explanation of them is contingent upon better understanding of the lower crust and upper mantle and the variations that may characterize the boundary between them.

A majority of the world's intracratonal basins have persisted through Phanerozoic time. Their activity, as documented by the details of their sedimentary fill, has been intermittent and of variable rate. Along with the regional arches that often form their margins, they are the dominant structural style of continental interiors. Their relative concentration of deeper-water sediments during periods of transgression, the great thicknesses of total deposits within them, and their considerable depth into the crust all combine to make such basins nearly ideal provinces for the generation of petroleum.

Hydrocarbons are often trapped by secondary faults and closures within these basins. ( Figure 2 , Various trapping possibilities associated with intracratonal basins and domes.

Figure 2

) 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 be consistent in orientation. ( Figure 3 , Structural contour map showing top of Precambrian, with four major faults. All of these have strong strike slip components of displacement.

Figure 3

Figure 4 , Regional cross section.

Figure 4

Note buried late Precambrian Keweenawan rift and Grenville front. Figure 5 , Maps showing basement provinces and the trend of productive anticlines in the overlying sedimentary basin fill. The overall NW-SE trend in fold strike is only suggested by the basement map.

Figure 5

The fight-hand figure shows locations of the wells productive in the deep Cambro-Ordovician as of mid-1984.)

A large number of traps in basins of this type, however, are stratigraphic and can be very subtle. During periods of active subsidence, downwarps act to localize certain depositional patterns. Unconformities; pinchouts due to sedimentary onlap; and reef, evaporite, and alluvial plain sedimentation are among the potential trap-generating characteristics that result.

Figure 3 , Figure 4 , and Figure 5 show the basic structural setting of the Michigan basin. The cross section indicates that, far from being a simple crustal sag, the floor of this basin is ruptured by the failed Keweenawan rift and the leading edge of the Grenville front, a probable megasuture of Late Precambrian age. Moreover, the "glove" of Michigan in general marks the enigmatic juncture of several principal shield provinces of North America.

Thus, it would seem that the crust in this specific location was potentially thinned and vulnerable to subsidence. Recent seismic surveys across Hudson Bay in Canada also show distinct basement block faulting. Though not a model for all intracratonal basins, this type of setting may indicate that deep-seated inhomogeneities in crustal character are an essential factor in the localization of such profound structures. Overall, simplicity of form disguises an apparent complexity of origin.

Other proposed ideas for intracratonal downwarping include the following: thermal contraction in the mantle; subcrustal flowage of material; mineral phase changes (causing local increases in crustal density); subcrustal erosion; differential cooling of the lithosphere. Crustal loading due to continued sedimentation is an obvious contributing factor, once, of course, subsidence has begun.

For exploration, the structural configuration of a specific basin will provide a broad context for locating potential traps within it (compare Figure 5 part (b) and ( Figure 6 , Distribution of principal oil and gas fields in Michigan.Traps are associated with several types of features: in the north and west, a Siluraian pannacle reef trend).

Figure 6

Rather than identifying structures only, seismic data may be of greater use in mapping certain productive intervals and their stratigraphic relationships. In Michigan, most of the older fields (now mostly in the stripper-well stage) were discovered in a trend of gentle anticlines in the central portion of the basin. The only giant producing area-the AlbionScipio trend (see Fig. 3 and Fig. 6) is associated with fracturing in Ordovician carbonates along a local strike slip fault. More recent production, however, has come from a Silurian pinnacle reef trend that rims the basin to the north and west. Though often prolific, these reservoirs are quite small and unrelated to structure, except in the most general terms of stratigraphic localization along the basin margin.

As a general rule, the structures of one intracratonal basin cannot be used as a specific guide to exploring other such provinces. Each basin is unique. For this reason, the explorationist is forced to become more of a specialist than might otherwise be the case. As an overall approach, petroleum geologists in Michigan have found it useful to extrapolate certain stratigraphic relationships observed near the margins of the basin into its center. This, however, will be risky in other provinces, given the facies changes that can occur. Thus, the range of traps shown in Figure 7 (Various trapping possibilities associated with intracratonal basins and domes) is only a guide.

Figure 7

Generally, however, with only a few exceptions, such as the Anadarko basin of central Oklahoma, most intracratonal basins have been explored only to relatively shallow levels. To some degree, therefore, a deep frontier remains in many heavily drilled areas. Recent gas discoveries below 15,000 ft (4500 m) in the Cambro- Ordovician sandstones of the Michigan basin and below 20,000 ft (6000 m) in the Ordovician Ellenberger group of the Permian basin in western Texas strongly support this idea.


Unconformities are primary structures whose identification and tracing are as important to structural geology as to stratigraphy. There are several reasons for this.

First, the period of erosion or nondeposition indicated by an unconformity marks a fundamental change in environment, one that is very often due to tectonic influence (this, however, must always be established).

Second, unconformities are invaluable markers for the deciphering of orogenic or epeirogenic events. Within a thick stratigraphic section, they divide the overall geologic history into a series of individual subhistories that can, to some degree, be analyzed separately. Mapping the distribution and characteristics of regional unconformities, in particular, is absolutely necessary to the understanding of geologic structure. For example, as shown in Figure 1 , reconstruction of paleoevents by the

stepwise removal of tectonic disturbance (sometimes called "pal inspastic reconstruction") is best done incrementally, by using each unconformity as a paleosurface.

Figure 1

Third, unconformities can be confused with certain types of faults, especially low-angle decollement thrusts through lithologies that have been subsequently intensely folded.

The principal types of unconformities are all seen on the cross section shown in Figure 2 (Cross section through Mills Ranch gas field, southern Anadarko basin, showing examples of an angular unconformity (A) several disconformities (D), and a nonconformity (N)).

Figure 2

The term disconformity is used where formations are parallel across a surface of nondeposition; angular unconformity is used when such formations display angular discordance; nonconformity is reserved for unconformities developed on crystalline igneous or metamorphic rocks.

The term hiatus refers to the total interval of geologic time that is unrepresented at a specific location along an unconformity. This almost always varies. Along regional unconformities, the differences can be as much as tens of millions of years. Hiatuses are most commonly measured by one or more of the following: qualitative geological time units (i.e., stages, epochs); biostratigraphy; paleomagnetic reversal correlations; and, where data for these are unavailable, lithostratigraphic correlations (e.g., individual members, formations, or groups). Examples of change in hiatus and how this is represented on stratigraphic cross sections are shown in Figure 3 .

Figure 3

This figure also shows how major unconformities divide a lithologic section into relatively separable sequences.

Recognizing Unconformities

Unconformities are often identifiable from surface, well log, core, and seismic information. Erosional truncation that involves discordance between units is very often nicely shown by seismic profiles. Most exploration, in fact, has come to rely heavily on seismic mapping of unconformities before drilling begins. At the same time, seismic data can only give us very general information about the thickness and specific character of an unconformity. Geologic description, therefore, is most often necessary.

Surface Recognition

Observation of outcrops remains the best means for analyzing the precise character of unconformities, once they are identified. Recognition, however, is not always straightforward, and the geologist must often be alert to clues in the form of the detailed effects of erosion. Angular truncation is one of the most obvious of criteria, and can be evident both in outcrop and in aerial photos. Discordance, however, is not by itself definitive evidence of an unconformity, as we shall discuss in a moment.

Marked changes in facies type or in the degree of induration across an identified contact are often good indicators of hiatus. Telltale signs of erosion, such as a basal conglomerate (containing pebbles of the underlying formation), chemical alteration (e.g., paleosoil profiles), discoloration, and irregular relief in the top of a formation, may also exist. Certainly, abrupt truncation of structures by relatively flat-lying units is diagnostic, as are significant differences in the intensity of deformation on either side of a contact. At the same time, however, some degree of structural variation may also result from normal ductility changes within a lithologic sequence, e.g., between a succession of dolomitized carbonates and a thick shale interval.

Paleontologic evidence can act as a deciding factor. Fossils sometimes indicate gaps of millions of years within lithologies showing no other obvious changes. In such cases, it must be shown that faunal successions are, in fact, interrupted, in order to guard against the influence of sample bias.

In certain areas, as mentioned, unconformities can be difficult to distinguish from faults. Such a case is shown in Figure 1 (Unconformity or fault? Continuity of a younger unit (Y) above an older unit (O) supports the first interpretation. This, however, is not conclusive.

Figure 1

). Where beds are tilted, perhaps even overturned, the true nature of the contact is impossible to discern by outcrop observation alone.

It is often helpful to remember, however, that with unconformities, younger beds will run parallel to the contact; thus the stratigraphic units above them should not be truncated. Nonetheless, decollement zones within a single, ductile lithology will not disturb original concordance for as much as tens of kilometers. Evidence for actual shearing due to movement may exist only within the immediate vicinity of the fault plane and in the sheared microscopic fabric of the rock. In all situations of doubt, evidence from as many sources as practicable should be considered before final determinations are made.

Subsurface Recognition

All criteria discussed regarding evidence for erosion also apply to examination of cores and cuttings. Narrow intervals of conglomerate do not always indicate paleoerosion zones; however, where such intervals separate well-indurated units of significantly different lithology, there is a good chance that an unconformity exists. Again, paleontologic work is often able to determine gaps in the rock record.

With sufficient supporting information, well log character can be used to trace the stratigraphic position of unconformities between wells. The detailed character of the unconformity and, especially, the lithologies on either side determine which specific logs will reveal the most identifiable response. As mentioned, sharp changes in rock type occur across many regional paleoerosion surfaces, and this will often influence an abrupt change in log character. Figure 1 ,

Figure 1

Figure 2 ,

Figure 2

and Figure 3 show examples of how gamma ray, sonic, SP, resistivity, and dipmeter logs respond to unconformities in several large productive fields.

Figure 3

In most cases where actual paleoerosion surfaces are present and identified in a section, the combination of one resistivity and one porosity log is sufficient to help locate and trace an unconformity. However, disconformities, especially those within single lithologies-most often produce no traceable change in log character.

In Figure 1 , note that the greatest change in log character takes place across the regional unconformity at the base of the Cretaceous (Long curves in the Ninian field, North Sea, showing multiple unconformities. Lithology is indicated by central column.). Here, lithology changes from dark, organic-rich shale to limestone. The difference is most conspicuous on the sonic log, which shows a sharp decrease in interval transit time from the shale into the carbonate, as we would expect. Likewise, as shown by the change in log response at the top of the Callovian, local unconformities can just as easily separate unrelated lithologies with very different petrophysical properties. In the case of angular unconformities, juxtaposed units continually change along the paleoerosion surface, and thus so does log character, as shown in Figure 2 (SP and resistivity logs through the Delhi field, northeastern Louisiana, showing multiple unconformities).

Figure 3 gives an example of how dipmeter logs can help identify unconformities (Resistivity and dip logs for well in northern Algeria, showing truncation of azimuth trends at unconformities). Abrupt discontinuity between dip azimuth trends is characteristic of angular unconformities. If trends are similar on both sides of the

unconformity, however, the change in dip amount, especially if slight, may appear to represent a depositional structure. At very low angles of discordance or across disconformities, no significant change may be visible. A weathered zone associated with a surface of paleoerosion may, however, be recorded as a zone of incoherent dips if a small correlation interval is used.

As mentioned, unconformities are very often identifiable on seismic profiles. This means that they can be traced, and their significance to some extent described, early in the exploration process. Discordance between lithologic sequences can be obvious, as we see in parts a and b of Figure 4 ,

Figure 4

or more subtle, as part c of Figure 4 shows (Seismic profiles showing examples of a nonconformity, b angular unconformity and c relatively subtle angular discordance decreasing to conformity, right to left ). In most cases, systematic termination of underlying reflections is strong evidence for either an unconformity or faulting. Figure 5 gives an example where terminations due to faulting and unconformities can be compared (Uninterpreted and interpreted seismic section from offshore western Africa, showing a complex diversity of both erosional and structural terminations in reflection patterns.

Figure 5

In bottom figure, the letters indicate ages of units: TPL=Tertiary, Pliocene; TM=Miocene; TP&E=Paleocene and Eocene; K=Cretaceous; J=Jurassic; TR=Triassic. Note especially the angular discordance between unit K1.1 (a deep-water clastic unit and underlying Jurassic strata; between units TP&E and K1.4, K2; and within the Tertiary section. In addition, an irregular unconformity separates the basal Jurassic from Triassic beds.).

Where the reflections that mark zones of discordance are continuous but show highly variable amplitudes, they are more likely to represent erosional unconformities. In many cases, surrounding structural and depositional patterns will help dictate whether faulting or paleoerosion has been responsible for the angular relationships observed.

Where lithologic change across a paleoerosion surface is substantial, and the associated velocity-density contrast is large, the unconformity may itself generate a reflection. This is true for both angular unconformities and disconformities. Often, the latter are identified by tracing out the former into deeper basin areas, as shown by part c of Figure 4 .


Certainly, folds are visually intriguing geologic structures. Geologists have long been drawn to the drama and spectacle of folded rocks. There is, as Hans Cloos (1939) observed, a special satisfaction and awe to be found in the evidence of earth stress having deformed originally planar and rigid rock into coherent arches many hundreds of feet high ( Figure 1 , Folded strata).

Figure 1

Even for the most seasoned structural geologist, the curving lines of folded strata never lose their particular aesthetic appeal. From a descriptive-analytic point of view, this degree of attention is justified by a pronounced variation in specific fold style, geometry, and generating mechanism, by attendant suites of complex but informative minor structures, and by the overall importance that fold structures have for petroleum exploration.

The specific attraction for geologists in folding lies in its direct implications concerning the origin, rate, evolution, and diversity of crustal movement. Several lines of evidence are considered in unraveling tectonic mechanisms:

the geometry of folds.

the experimental simulation of folding.

the specific relationship of folding to faulting.

the relationship of fold belts to other tectonic provinces (analysis of structural style).

the small-scale structures associated with folding.

To some degree, modern petroleum exploration was born with the anticlinal theory. I.C. White's formulation of this hypothesis in 1885 represents the first conceptual approach to systematic and successful prediction of oil and gas occurrence. The elegance of this theory lies in its simplicity and in its enduring practical application, which, with the aid of modification and expansion, continues today. The anticlinal theory, however, is not the only direct benefit that exploration and production geologists have derived from a knowledge of folding. A few others include:

the prediction of fracture trends.

the prediction of fracture distribution and intensity.

the explanation and prediction of tectonically induced variations in reservoir thickness, porosity, pressure, and general fluid behavior.

the prediction of oilfield shape and extent.

the delineation of constraints on contour and isopach work.

the prediction of deeper accumulations beneath known fields.

the understanding and prediction of significant faulting and its influence.

the understanding and prediction of the general orientation and style of neighboring structures, and, therefore, the location of potential traps in other parts of a specific region.

Description of Folds

A fold is produced when initially planar layers become bent, i.e., nonplanar. To the naked eye, folds in rock appear to be continuous; layering is preserved, not truncated. For many years, geologists took the coherent curvilinear geometry of folds in sedimentary strata to mean that strain had accumulated gradually and continuously. Within the past seventy-five years, however, it has become evident that a great deal of fold deformation is accomplished by discontinuous, incremental gliding along bedding planes, within layers, and between and inside of individual grains and crystals. Thus, the strain involved in folding imposes change in both the geometry and internal physical properties of the layers, which normally remain relatively coherent.

Folding is seldom the result of a single, discrete deformational episode.

Diastrophism nearly always occurs in what can be described as pulses: during the evolution of an orogen, stages of intense tectonic activity wax, wane, and frequently overlap at irregular or semi-regular periods that can affect different parts of a specific region at different times. The reasons for this have not yet become clear, but they are assumed to be more closely tied to the nature of stress generation within the earth's crust, rather than to the mechanics of strain response.

The most basic types of folds are the anticline, in which right-side-up bedding dips away from the fold crest, and the syncline, in which bedding dips inward. Regional arching with associated folding produces an anticlinorium. Similarly, large-scale downwarping of a fold belt creates a synclinorium. ( Figure 1 , Example of anticlinorium and synclinorium from the Rhineland region of Germany.

Figure 1

Note the large scale of structures. Dashed lines indicate cleavage.) Generally, the orientation of the anticlinorium or synclinorium roughly parallels that of the folds within it.

The terms anticline and syncline can only be used once the younging direction is known. In strongly folded regions, beds may be completely overturned. Thus, during the early stages of analysis, before stratigraphic relationships have been established, it is common practice among structural geologists to make use of the terms anti form and synform, neither of which implies the direction of younging.

To describe the great variety in fold morphology, a correspondingly diverse nomenclature exists. This discussion uses only more fundamental and commonly employed terms. The reader who is interested in more detailed study is referred to the extensive treatments given by Whitten (1966), Ramsay (1967), and Davis (1984).

Basic Fold Geometry and Orientation

Figure 1 displays the principal descriptive features of a simple fold.

Figure 1

Several of these represent imaginary lines and surfaces that help us analyze a fold as a geometric phenomenon. Except for those that require the third dimension for their depiction, these features are most often shown on cross sections drawn perpendicular to the long direction of the fold. This type of section is called a fold profile. In Figure 1 , the definition of each term should be fairly obvious from the geometry shown.

The hinge of a fold is its point of maximum curvature at one particular location (cross section) along its length. The fold axis, or hinge line, contains all hinge points in a folded layer. The axial surface, then, connects all these axes within a single fold composed of multiple layers. In some cases, the axial surface is planar and can be called the axial plane. The plunge of a fold is the angle between its axis and a horizontal plane. ( Figure 2 , Illustration of fold plunge and its change along strike. Such culminations and depressions characterize folds of many scales in mountain belts.)

Figure 2

For the purposes of practical simplification, geologists often treat folds as if they were cylindrical, i.e., generated by a line moved parallel to itself in space ( Figure 3 , Perfect cylindrical fold system), or conical, i.e., generated by the same line with one end fixed.

Figure 3

Noncylindrical and nonconical folds are common in nature, but can often be considered the summation of cylindrical or conical parts ( Figure 4 , Complex folding

in the Vanige region of southwestern France).

Figure 4

As a result, for extrapolation to greater or shallower depths, geologists use the interlimb angle to project fold geometry ( Figure 5 , Classification of folds on basis of interlimb angle. Lower figure indicates how this angle is determined for curvilinear folds).

Figure 5

In petroleum geology, folds are most often drawn in vertical cross section (which is different from a fold profile if the structure is plunging) and are shown as cylindrical, even to depth. This must be recognized as an approximation, especially in areas where thick shale or evaporite intervals exist.

In profile, folds are commonly described as symmetrical or asymmetrical on the basis of whether their axial surface is a plane of symmetry or not. More specifically, as shown in Figure 6 , folds can be classified as upright, overturned, or recumbent on the basis of the inclination in their axial planes.

Figure 6

Large-scale recumbent fold structures, generally accompanied by thrust faulting, are often called nappes (German "decke"; English "sheet" or "cover"), a term originated by Alpine geologists to describe large, thin, flat-lying structures that have been transported like rumpled sheets for considerable distances ( Figure 6 and Figure 7 ).

Figure 7

This style of deformation in unmetamorphosed rocks is generally unusual except in the European Alps.

Geometrically, one of the simplest forms of folding is the monocline. ( Figure 8 , Simple block diagram of monocline.

Figure 8

Note how upper layers are passively draped over faulted basement.) This is frequently a steplike drape in sedimentary cover over faulting in crystalline basement rocks. It basically represents, therefore, a form of passive bending, in contrast to the active buckling that produces most anticlines and synclines in orogenic zones.

Fold Closure

The principal importance of folds in petroleum geology is very often related to the concept of structural closure. As shown in Figure 9 , (Definition of structural closure.

Figure 9

Part a shows that it is the crest of the fold that determines closure (not the hinge); b illustrates how the geometry of closure varies with that of the fold.) closure is the pocket formed between the crest of the fold and the lowest closed contour. It is measured in terms of the vertical height of this pocket and is an overall indicator of a fold's total capacity to hold oil and gas, given good reservoir properties. Structural closure should be contrasted with structural relief, which is the relative height to which the fold rises above the regional slope.

Basic Fold Classification Parallel and Similar Geometry

In addition to describing folds according to their geometry and orientation, geologists have found it useful to classify them on the basis of morphological criteria that reflect mechanisms of deformation. This classification scheme examines folds in profile. Its principal attempt is to distinguish structures that have formed by flexure and layer-parallel shearing from those that have resulted from flow. Figure 10 shows two idealized profiles of these types.

Figure 10

Parallel (also called "concentric") folds are so named because the upper and lower surfaces of individual layers remain parallel during folding. Note in the figure that this does not mean all layers necessarily remain parallel to each other. It does mean, however, that they maintain a constant thickness throughout the fold (part a of Figure 10 ). This makes for crowding of hinge areas, which either causes the fold to die out rapidly with depth or to compensate the increasing lack of space by faulting ( Figure 11 , Hypothetical cross section to illustrate how parallel folding creates overcrowding and faulting in hinge areas.

Figure 11

Note also the extensional fracturing along the crest of the fold) and basal detachment (also called "decollement") ( Figure 12 , More detailed prediction model based on observed structures in productive areas of the Rocky Mountain region of Colorado and Wyoming).

Figure 12

In Figure 10 , this crowding is shown as an increased crumpling in the thinner layers. This is one form of what is termed disharmonic folding, so called because of the difference in wavelength.

Another way to view this hinge crowding is to note that with depth in a parallel fold, anticlines decrease in size, while synclines increase. Generally speaking, parallel folding is more common when deformation is of moderate or low intensity and involves thick, competent units.

In similar folding, by contrast, substantial flowage of material is assumed to take place. This can occur in two basic ways: (1) by material transfer out of the fold limbs and into hinge areas, and (2) by movement along planes of cleavage that develop parallel to the axial plane of a fold.

Geometrically, the shape and size of a similar fold is retained at depth (part b of Figure 10 ) without faulting. The flowage required for similar folding is most common in incompetent lithologies (e.g., shales and evaporites) and at elevated temperatures and pressures. Slippage along axial plane cleavage, as we will see, often leads to similar folds in shales.

Though both parallel and similar end-member types are seen in the field (similar folds being particularly characteristic of metamorphic rocks), folds in sedimentary

rocks more often show a compromise between them. For example, folds that have concentric limbs, often show thickening at their hinges. Particularly in stratigraphic sequences with high ductility contrast for example, competent sandstone or carbonate units separated by thick shale or salt intervals-substantial flow from sheared limbs into hinge areas should be expected.

In such cases, competent layers show more or less parallel folding. This fact is often important to keep in mind when structures are projected to depth in cross sections. Even a moderate degree of similar folding will prove depth predictions based on parallel geometry to be in substantial error.

In general, folds formed under low confining pressures-such as most of those that trap hydrocarbons-are mainly the result of rigid body rotation and layer-parallel shearing. Such a mechanism produces parallel folds in competent layers, but may be accompanied by flow in incompetent lithologies. Figure 13 and Figure 14 give two examples of such folds.

Figure 13

(Figure 13 is a cross section through Spring Creek field in northwest Wyoming (western margin of the Big Horn basin), showing many of the structural features predicted in Figure 12.

Figure 14

Note how closure changes with depth due to faulting. Figure 14 is a classic section through the Jura Mountains of eastern France showing sharp, boxlike folds developed above a shallow decollement plane. Note how these folds show a combination of parallel and similar geometry, as well as pronounced flowage of Triassic gypsiferous marls into hinge areas. )

Fold Mechanisms and the Distribution of Strain

The distinction between parallel and similar folding is very useful as a general guide to fold geometry in many petroleum provinces with relatively simple structural traps. However, in more complex regions, where rocks have developed secondary tectonic features (e.g., cleavage or fractures) that can influence entrapment, it is sometimes necessary to determine mechanisms of folding in order to predict the potential existence and geometry of traps. Fracturing is often genetically related to folding; therefore an understanding of mechanisms will to some degree also explain the existence, orientation, and trends of fracturing.

Geologists working mostly in regions of sedimentary rock emphasize the bending or buckling of layers, while those more familiar with metamorphic terrains have stressed the ability of rocks to flow. The great diversity of fold structures in the earth has resulted in the development of competing terms and concepts. Here, we attempt to simplify these and make them easier to understand by looking first at mechanisms that describe how individual layers become folded, and then at mechanisms that relate to the folding of multiple layers.

To explain what is seen in nature, we can identify three main processes that seem to explain the folds we might observe in single layers of any lithology. The principal

importance of these processes for our understanding of folding is that they describe what happens within a particular layer.

1. Pure bending (part b of Figure 1 ) occurs by the simultaneous stretching of material along the outer arc of the folded layer and the compression of material along the inner arc.

Figure 1

Between these zones of relative tension and compression is a "neutral surface" where no strain occurs. The compensation of stretching by shortening means that the resulting fold will be parallel in its geometry.

2. Flexural shear (part c of Figure 1 ) deforms a layer by internal shear along planes that are parallel to the folding layer. As a whole, the layer can be thought of as undergoing a shear couple, with the relative sense shown in part a of Figure 1 .

Note that with this mechanism, no deformation (such as stretching) occurs along the plane of the fold layer. Slip takes place along discrete surfaces within the layer.

A special type of flexural shear folding is termed flexural flow. This is said to occur when deformation takes place on the scale of individual grains, and there are no discrete surfaces of displacement. A certain amount of flow into hinge areas often

occurs during this type of folding. This is generally small and depends on the lithology involved.

Flexural shear produces folds with mostly parallel geometry.

3. Shear folding (part d of Figure 1 ) occurs when shearing takes place within a layer along closely spaced planes oriented obliquely to the layer itself. These planes are often roughly perpendicular to bedding in hinge areas and oblique to it along fold limbs. This mechanism involves simple shear and produces similar folds. It is much more common in ductile lithologies and in medium- or high-grade metamorphic rocks. In thick shale intervals, shear folding is apparently a late-stage mechanism that begins after the development of penetrative axial plane cleavage during flexural shear.

Each of these mechanisms produces plane strain in single layers. Each, admittedly, represents an ideal condition and is only occasionally seen in nature in its "pure" state. This is because the process of folding is strongly determined by the character of the entire stratigraphic sequence involved. Through the study of laboratory models, geologists have confirmed their intuition that folding is often the result of progressive mechanical adjustment both within and between individual layers.

In many cases, it appears that single layers can shorten (through compaction, among other processes) by as much as 20% before folding actually begins. This means that a certain amount of strain hardening may occur very early. At the same time, most stratigraphic sequences are characterized by high ductility contrast, with sandstone and carbonate units frequently being the more competent, and shale or salt the less competent, intervals. Generally, it is the former that fold first and control the subsequent deformation to a large extent. Response of less competent layers, therefore, is relatively passive.

With regard to the folding of multilayered sequences, it is important to be familiar with three basic mechanisms. These are the following:

1. Flexural slip refers to the shear-or slip-that occurs between layers as a fold develops. As implied by part a of Figure 2 , this slip compensates the bending and rotation of comparatively rigid layers that do not undergo large amounts of internal plastic deformation.

Figure 2

It therefore occurs principally while a fold is still open. Thinner, more ductile layers between competent layers are subjected to relatively intense and penetrative shearing. Notice that this style of folding requires that tectonic compression occur parallel or subparallel to bedding. Looking at the mean strain ellipses in part a of Figure 2 , we might note how similar are their orientations to ellipses in the limbs of a fold developed by flexural shear (part c of Figure 1 ). Flexural slip folds are the most common folds in nonmetamorphosed sedimentary rocks.

2. Shear folding is a topic we discussed with respect to the internal deformation of single layers. In multilayered sequences, shear planes (such as cleavage) can cut across competent layers and thus allow for a certain (usually minor) amount of slip (see part b of Figure 2 ). Such folding, obviously, would succeed earlier deformation that created the shear planes. In general, shear folding becomes significant at higher grades of metamorphism.

3. Kink folding (part c of Figure 2 ), also called chevron folding, is caused by a combination of flexural slip between layers and sharp, localized yielding (with some ductile flow) in hinge areas. Kinking is most often seen on a mesoscopic (hand-sample) scale in the field and commonly occurs as a conjugate set of single kink bands. (The term "chevron fold" is most often reserved for larger structures, with amplitudes measurable in meters rather than centimeters.) The details of kink band formation are somehow linked to the anisotropy present in thin-bedded sequences of alternating competent and incompetent

layers. As a general rule, kinking is of relatively minor importance to petroleum exploration, though a few important exceptions exist, such as the Ventura anticline in southern California, with its reserves of nearly a billion barrels.

One final component of the folding process should be mentioned. In addition to the various mechanisms we have discussed, many folds also undergo a component of flattening. This additional shortening, shown in Figure 3 , takes place approximately normal to 1 with simultaneous extension in the 3 plane.

Figure 3

As the figure shows, sufficient flattening may "tighten" a fold by pure shear strain bringing all strain ellipses into relative parallelism and originally flexural slip folds into more or less similar fold geometry.

Some geologists believe that almost all folding involves this type of additional strain (see Ramsay 1967). Relevant experimental and theoretical work has, thus far, been limited, but has apparently confirmed the assumption that flattening is more important in sequences of low ductility contrast.

Given the an isotropy and consequent ductility variation in most lithologic sections, the folding process takes place by sometimes complex combinations of these end-member mechanisms. This we should expect. A relatively competent siltstone layer, for example, may deform internally by a degree of flexural shear and "externally" by flexural slip. A very general rule of thumb predicts that with increasing depth, and thus ductility, a given layer will show transition from parallel to similar folding.

For the petroleum explorationist, knowledge of what might be called the structural character of the stratigraphy in a specific area is essential. Such knowledge will act as a basis for developing an intuitive sense of what types and geometries of structure may be expected.

Moreover, different fold mechanisms have different effects on the petrophysical character of rock. The greater amount of internal shear a sandstone unit has suffered, for example, the more its porosity-and possibly permeability-will be directly related to position on a fold.

More generally, tectonic compaction can substantially reduce porosity along fold limbs and in hinge areas; pressure solution in carbonates can cause considerable permeability loss. At the same time, the flattening or stretching of clastic grains can create directional permeabilities. Pressure gradients produced during folding can also cause fluid migration, with simultaneous dissolution of material and porosity enhancement. Fracturing can occur at many stages of the folding process. If interconnected, fractures can result in tremendous amplification of reservoir quality. These are a few of the main effects that folding can have on prospective reservoir formations.

As we have mentioned, most exploration takes place at relatively shallow structural levels in the earth. Thus, competent units (sandstones, thicker carbonates) are usually assumed to deform into parallel folds (see Figure 5 and Figure 6 ).

Figure 5

Particularly when such units are relatively thick, the extra space created between them in fold hinge areas is assumed to be filled by intervening ductile material, especially shale, marl, and, where available, salt or gypsum.

Figure 6

The strong component of flexural slip between competent units is thought to be the major influence forcing weaker material out of fold limbs and into hinge areas. The actual amount of material displacement is rarely known; resulting cross sections, therefore, very often portray folds as parallel, with near-ideal geometry. This is a practical simplification. However, if the stratigraphic section is well known, folds can be drawn to reflect the flowage of material that we would expect to occur in the most ductile intervals. This is often necessary in order to accurately project fold geometry-and thus closure-to depth.

Pronounced ductility contrasts between relatively thin, brittle layers and thicker, incompetent units often result in disharmonic folding (see part a of Figure 4 ). Here, the wavelength of folds varies considerably between layers. As mentioned above, disharmonic folding is a frequent and predictable occurrence in the hinge areas of concentric folds, due to the increased crowding that occurs with depth.

Figure 4

In certain orogenic provinces (especially fold-thrust belts), this type of deformation is very common on both a mesoscopic and macroscopic scale. Comparatively thin sandstone or carbonate layers (1-2 m) within thick shales (50-100 m) are often intensely crumpled within large, open folds that have wavelengths of a kilometer or more. Such layers thus indicate directly the amount of flow that has occurred. They are, however, local features, and should not be mistaken as signs of the general intensity or style of deformation.

In general, overcrowding in the hinge areas of flexural folds can greatly disrupt layering by disharmonic folding, fracturing, and minor faulting. This will be more true for folds with interlimb angles less than about 40-50º. Within such a highly deformed hinge zone, dipmeter data may show a relatively sudden change from a continuous pattern to a random, "bag o' nails" motif.

With respect to potential reservoir formations, the concentration of hinge strain can have several effects. It can, for example, enhance porosity through fracturing, or can destroy it by tectonically compacting the stronger, more resistant grains. Such strain may also cause the release and mobilization of solutions that will dissolve soluble material out of the rock matrix and recrystallize it in fractures, thus sealing them completely. Carbonate material, in a calcareous sandstone, for example, is particularly susceptible to this type of local mobilization.

Minor Structures Associated with Folding

A great variety of minor, secondary structures generated during folding have been identified by geologists working in the field and subsurface. These features can be

extremely useful for the unraveling of complex fold geometries. Wilson (1982) offers a comprehensive listing of them. Only a few of the most common are relevant to subsurface exploration and thus are included in the discussion below.

Kinematic Axes and the Description of Minor Structures

Figure 1 indicates what structural geologists call the kinematic ("motion-describing") axes of a flexural slip fold.

Figure 1

These axes are generally used to indicate what a particular minor structure reveals about the geometry of the host fold. As shown, the orientation of the three axes changes across the fold. Note that the a-axis represents the direction of slip between layers (called "S surfaces"); the b-axis is parallel to the axis of the host fold; and the c-axis is the direction perpendicular to (across) layering.

The a-axis is sometimes referred to as the direction of tectonic transport, with the a-c plane describing the plane in which most deformation (folding, flattening, stretching) takes place. Interlayer slip occurs along the a-b plane.

The general term "S surface" is used instead of "bedding" for two reasons: (1) slip sometimes occurs within single layers, thus paralleling but not necessarily coinciding with bedding, and (2) in highly deformed rocks, bedding is not always obvious; secondary planar fabrics (S1, S2, and so forth) may exist and need to be distinguished from true primary layering (S0)

Where cleavage is well developed, the kinematic axes are oriented as shown in Figure 2 .

Figure 2

To some degree, this is based on the assumption-usually justified-that a significant portion of later deformation in such folds occurs by shear along cleavage planes.

Figure 3 shows a variety of common minor structures, including joints, minor folds, cleavage, and lineations, that can be used to derive the geometry of the host fold.

Figure 3

Among the most important of these to identify and measure are b-axis lineations, i.e., those that parallel the main fold axis. For example, in structurally complex areas, the intersection of bedding and cleavage can sometimes be seen and measured in oriented core samples. Elongated pebbles, flattened ooids, and deformed calcite and quartz grains can all be interpreted in terms of kinematic axes.

Minor Folds

Figure 4 shows the assumed origin of what are known as parasitic folds.

Figure 4

This is also sometimes called "drag folding," but the more common use of the term drag folding in the petroleum industry is in reference to folds presumably caused by faulting. Parasitic folds frequently characterize disharmonic layers within flexural slip folds. Shearing between thicker, competent intervals is taken up by the development of these structures. Their sense of shear indicates which limb of the host fold they occupy, and their axes generally approximate that of the host fold.

Cleavage

This is a complex phenomenon of deformed sedimentary rocks that can be basically described as a tectonically induced, planar fabric that imparts a mechanical anisotropy to a rock. A comprehensive review of the mechanisms proposed by various researchers to explain the origin of cleavage is given by Wood (1974). For our purposes, two basic types deserve mention.

Slaty cleavage, also called "continuous" or "flow" cleavage, results from the parallel alignment of all platy (clay and mica) minerals perpendicular to the direction of maximum shortening. It apparently occurs as the combined result of de-watering, grain rotation under stress, and a certain amount of recrystallization in response to flattening.

Slaty cleavage exists throughout a specified material and completely dominates its mechanical properties: it is therefore called "penetrative." Where especially well-developed, it accompanies the low-grade metamorphism of shale into slate. A good slate, such as that from the quarries of North Wales, could be split with the ideal tools

into sheets as thin as-or even thinner than-a single sheet of paper. In folds, slaty cleavage develops parallel or subparallel to the axial plane, and is therefore often referred to as an "axial plane foliation" (see Figure 2 and Figure 5 , Slaty cleavage in a small, overturned fold, Canadian Rockies).

Figure 5

It can also, however, develop as a result of intense shearing between thick, competent units deformed by flexural slip, or in association with faults. The precise stage of folding at which slaty cleavage is formed appears to be variable; however, it does seem to be a relatively early development. Many structures show a fanning in the orientation of cleavage planes that seems to be caused by continued folding after cleavage formation ( Figure 6 , Progressive development of slaty cleavage, showing late-stage minor shearing along cleavage planes. Note the thickening of individual layers as a result of shortening).

Figure 6

Fracture cleavage, also called spaced cleavage, is commonly a form of closely spaced jointing, and results from the development of discrete planes of mechanical discontinuity within a relatively competent bed ( Figure 7 , Diagram showing the proposed origin of fracture cleavage due to intralayer stress generated by flexural shear folding.

Figure 7

Normally S1 fracture plane develops; S2 does not. Continued shearing will rotate fractures hingeward.). Generally, no grain reorientation is involved. Actual parting may be incipient or fully developed, and movement along the planes of fracture (which transforms such cleavages into "microfaults") may also have occurred ( Figure 8 , Sketch of observed fractures in sandstone layer enclosed by cleaved shale, Delaware Watergap, eastern Pennsylvania).

Figure 8

Fracture cleavage is thought to be the result of shearing within competent units deformed by flexural slip (Billings 1972; Dennis 1972; Hobbs, Means, and Williams 1976; Park 1983). Its spacing is related to how brittlely a layer behaved during deformation: massive, competent units will show wide spacing; weaker, thinner layers may show as much as 150 fracture planes within a single centimeter (Wilson 1982). Figure 9 shows how the attitude in cleavage planes can vary with lithology.

Figure 9

Within a single fold, this change across layers is usually termed refraction of cleavage.

As a form of fracturing in competent units, this type of cleavage can have obvious significance for petroleum accumulation and migration. Once its geometric relationship to bedding is accurately determined, its orientation can be predicted on the basis of position in a fold. Inversely, it can itself be used to indicate what portion of a fold is being examined.

Fractures

In addition to cleavage, discrete sets of fractures often characterize folded competent layers. Stearns (1967,1977) has identified five major fracture patterns useful for a detailed understanding of brittle behavior in fold deformation. The common occurrence and interconnected nature shown by two of these patterns makes them important to understand. Their consistent relationship to bedding attitude indicates that they are, in fact, the result of folding

Pressure Solution and Stylolitization

In addition to cleavage, certain changes in rock fabric occur as a result of tectonic stress. Two of these, pressure solution and stylolitization, can have especially large effects on rock porosity and permeability. Though both are found in association with faults as well as folds, they appear to be more common in highly folded rocks that have transmitted and absorbed stress internally.

Pressure Solution

This is a phenomenon that involves the dissolution of material at points where there is a concentration of stress, and the reprecipitation of that material in places of lower stress concentration. The process of pressure solution is usually explained in terms of Riecke's principle, which states that (1) material will dissolve under pressure on the sides of objects (single grains, crystals, pebbles) that face the principal compressive stress, and (2) this dissolved material will reprecipitate on those sides facing the least principal stress.

Basically, solution occurs because the tectonically induced pressure exceeds the hydraulic pressure of the interstitial fluid. It is most common along grain, crystal, or pebble contacts, and its main physical effect is to reduce pore space and more tightly weld a rock together. If extensive, pressure solution may cause consistent elongation of individual grains or cobbles. Where the original shape of these is known or can be determined, the principal axes of stress can be derived.

Pressure solution is a very common feature of folded sandstones and carbonate rocks, especially limestones. Due to the relatively high water solubility of calcite, material goes into solution quite readily during intense folding. Quartz grains and pebbles also frequently show evidence of pressure solution, even in rocks that have been only gently folded. Surrounding conditions, such as pressure, temperature, and the composition of pore fluids, naturally have a strong influence on the degree of pressure solution that results.

Stylolitization

This is a special case of pressure solution that occurs predominantly in carbonate rocks to produce thin, sawtoothlike seams or contact surfaces. In cross section, these seams resemble the irregular tracing of a stylus. It is said to mark a zone of solution resulting from differential vertical movement that may or may not be related to the folding process. Some stylolites, for example, are parallel to bedding and have apparently originated during diagenesis.

Stylolites are especially common in relatively homogeneous carbonate lithologies. Dissolution involves the removal of the most soluble (carbonate) material from the seam, which is most often filled by residual insolubles such as clay, sand, iron oxides, and organic material. This sometimes allows their presence to be detected on natural gamma ray logs as peaks within otherwise relatively unradioactive carbonate units.

Stylolites are known to be permeability barriers. Where common, laterally extensive, and parallel to bedding, they can significantly influence the patterns of fluid flow within a prospective reservoir. The extent of individual stylolites is generally unpredictable, and can range up to hundreds of meters. More often, however, they are relatively local. They frequently mark boundaries between similar lithologies of different texture, but can terminate, begin again, overlap, or bifurcate unpredictably.


Like faults, joints result from a brittle response in rocks to stress. They are the most conspicuous and omnipresent secondary structure of rocks exposed at the earth's surface. In subsurface work, petroleum geologists use the term "fractures" to refer to local ruptures of almost any kind that do not show enough offset to be called faults. Geologists in the field distinguish joints from faults in a similar fashion, i.e., on the sole criterion that no visible displacement has occurred along the planes of parting. Ironically, the term "joint" was coined over two hundred years ago by workers in British stone quarries. These men believed that the pleasingly regular, mutually perpendicular planes were those across which individual blocks of rock were joined.

Most often, joints occur in sets of semi-regular spacing. They may be fully developed and conspicuous or incipient ( Figure 1 , Illustration of joint face structures and nomenclature).

Figure 1

Frequently related to regional deformation patterns, joints are undoubtedly the source of reservoir fracturing in many areas. They may also result from non-orogenic stresses initiated by draping and differential compaction.

Joints are generally classified by nomenclature that reflects particular mechanisms of formation. Extension joints (a subset of extension fractures) develop perpendicular to 3. Sheeting or exfoliation joints are basically a type of extension jointing that develops parallel to the surface of the earth. Such fractures are presumably caused, at least partially, by the release of confining pressure in a fairly isotropic rock body (e.g., an igneous intrusion) as its overburden is decreased through erosion. Shear joints, as a form of shear fractures, develop at acute angles to 1.

Within a particular set, joints need not be parallel. Typically, jointing orientation is related to position on a particular fold. Moreover, two or more sets frequently occur together, comprising a joint system that essentially splits a specific rock body into an assembled mosaic of blocks.

Microfractures, which require microscopic examination for their identification, are more common in some competent lithologies. On the other hand, visible joint sets in sandstones or dolomites can show wide spacing (tens of meters) and can exert a dominant control on surface topography.

Fracturing is a relatively shallow structural phenomenon in the earth's crust and is highly dependent on lithology. It remains unclear at what depths joints can be generated. On the basis of mathematical considerations, tension fracturing is predicted to a maximum of about 3 km (Hobbs, Means, and Williams 1976). However, given a sufficiently low geothermal gradient, high pore pressures, and a large differential stress (1 -3), both extension and shear fractures can be produced at considerably greater depths.

Types of Fractures and the Influence of Lithology

As a general rule, factors that tend to increase rock ductility decrease the overall contribution of rupture to deformation. In a stratigraphic sequence of high ductility contrast, its spacing and orientation can vary substantially between major rock types. To help quantify and better analyze this variation, Stearns (1967) has derived the concept of "fracture number." This he defines simply as the average number of parallel fractures per 100 ft (33 m) normal to the fracture plane. For this type of measure, any linear distance could, of course, be used. This, then, makes fracture number a potentially useful quantity for core and microscopic analysis. Average fracture numbers for several competent lithologies are shown in Figure 1 .

Figure 1

According to Stearns and Friedman (1972), fracture systems that show consistent orientations and that pervade a large volume of rock can be divided into two broad types: (1) those related to specific structures, and (2) regional orthogonal fractures. The former type can most often be explained in terms of the stress system that created the host structure. The latter type, however, is not well understood. Some geologists have related it to epeirogenic movements, most notably plateau uplift, but this does not appear to explain most cases of regional fracturing. Both types of fractures, however, can enhance the reservoir quality of prospective formations.

Fracturing can increase the effective permeability of rock by as much as several orders of magnitude. It is responsible for the creation of reservoirs in rocks that normally lack sufficient porosity and permeability to hold hydrocarbons, e.g., shales, quartzites, and even igneous and metamorphic lithologies. Generally, this permeability increase is greater along extension fractures, since these often undergo a small amount of separation.

By contrast, there is a component of compression along shear fracture planes. This means that certain preliminary assumptions about directional permeability can be made in cases where fractures can be tied to a specific stress system. Usually, this involves an analysis of the relationship between fracturing and known structures, such as faults and folds.

Basic Techniques for Determining Fracture Orientation

Any determination of fracture orientation should be based on statistical analysis. Several techniques exist and examples of each for a single joint system are shown in Figure 1 (Part a is a rose diagram, part b is a strike histogram and part c is a stereogram, showing poles to fracture planes.).

Figure 1

In cases where fractures are mostly vertical or near-vertical (which is usually the case), rose diagrams and histograms showing strike frequency are used.

Rose diagrams are often preferred, since they can be plotted directly on maps to show the actual orientations and relative dominance of different fracture trends ( Figure 2 , Major fracture patterns in uplift areas of the Colorado Plateau.

Figure 2

Note the general NW-SE and NE-SW trends. These have been interpreted by some researchers as indicating conjugate failure on a regional scale.). However, use of stereo-grams is essential in cases where fractures dip at substantially less than 90º. Stereograms also allow fracture and fault or fold data to be plotted and compared on the same figure. This is extremely useful for establishing structural relationships in fracture analysis. Notice that in part c of Figure 1 , fracture set Ill closely parallels the host fold axis (point B). As a preliminary assumption, we might consider this set to represent extensional fractures; sets I and II, therefore, most likely indicate conjugate shear fractures.

Fault-Associated Fractures

Faults and fractures both represent stress-induced rupture of rock, and that when they occur in association they can generally be related to the same stress field.

Fault movement itself generates shearing stresses that can induce fracturing. In many instances, therefore, shear fractures can be considered miniature versions of a particular fault. Thus, knowing the orientation of a fault means that one can sometimes predict associated fracture trends. This also means that fracture orientation will change with fault attitude.

As a general rule of thumb, faults that develop at shallow levels in especially competent lithologies (e.g., dolomite) are more likely to have fractures associated with them. However, it should be emphasized that no necessary relationship exists between the displacement along a fault and the amount or intensity of fracturing.

Shear fractures are more likely to undergo displacement when associated with faulting than with folding. Such movement can either increase permeability by creating a poor fit between the two sides of the fracture, or decrease it by sealing the fractures with gouge or even mylonite. As a rule of thumb, the intensity of fracturing can be expected to be relatively equal in both upthrown and downthrown blocks of a normal fault, but somewhat higher in the upthrown block of a reverse or thrust fault.

According to Stearns and Friedman (1972), several basic rules can be applied to drilling a well such that the greatest number of natural fractures are encountered. As shown in Figure 1 (Principal fracture patterns and their respective strain ellipses associated with normal and reverse faults.

Figure 1

In each case, there are three possible fracture sets: two conjugate shear fractures and one extensional fracture. One conjugate parallels the fault; the other is antithetic to it. Note that the strike of fractures theoretically parallels that of the host faults or faults.), the strikes of all three potential fractures will generally parallel that of the host fault. For low-dipping faults, no deflection of the borehole is needed to intercept the greatest number of fractures. As the fault attitude steepens, deflection of a well

toward the fault plane becomes more necessary ( Figure 2 , Diagram showing the dependence of fracture orientation on fault attitude. N refers to normal fault, R to reverse fault. Angles between shear conjugates are idealized.).

Figure 2

Fractures can develop during the early stages of deformation and may thus become rotated. Later folding or faulting of a fractured section may obscure original structural relationships. Most orogenic regions experience multiple episodes of deformation; thus, later trends often overprint earlier structures. At times, the true connection between fracturing and faulting will become clear only after these later deformational effects have been "removed

Fold-Associated Fractures

Other sets of fractures besides cleavage often characterize folded competent layers. As an example, Figure 1 shows the variety of fractures seen in a small anticline in southern Germany.

Figure 1

The fold shows an early stage of cleavage formation in finer-grained lithologies, and various shear and extension fractures in the thicker, more competent sandstone layers.

In all, five fold-related fracture patterns have been identified and analyzed (see Stearns 1967). Figure 2 shows the derived axes of greatest and least principal stresses for the two most common patterns (Patterns associated with folding (mostly parallel); note that both patterns show a consistent geometric relation to bedding.).

Figure 2

As with faulting, these patterns consist of conjugate shear fractures, plus an extension fracture. Pattern A indicates stretching along strike, parallel or subparallel to the host fold axis. This has been documented in folds of many styles and scales and is frequently indicated by b-axis lineations, such as grain or cobble elongation (see Figure 3 , Various common minor structures useful for analysis of fold geometry.

Figure 3

Shown are b-axis lineations and elongated cobbles; a-axis striations due to interlayer slip; minor folds with axes parallel to that of the host fold (B); a-c joints; and cleavage planes.). Pattern B is essentially the same set of fractures rotated 90º, and indicates that stretching takes place in the plane of dip. This pattern should recall the distribution of strain shown in part b of Figure 4 (pure buckling).

Figure 4

Fracture cleavage, can be thought of as a shear fracture pattern.

According to Stearns and Friedman (1972), both patterns A and B can characterize a single bed. Individual shear fractures of pattern A often occur as relatively isolated features. They can cross an entire fold and extend hundreds of feet vertically, but are also seen on many scales, even that of single grains. They show exceptional consistency in their orientation on all scales, however, which means that statistical plots show nearly identical patterns, whether data are taken from aerial photographs or thin sections. Pattern B fractures are smaller (up to several meters long), but all three fracture sets usually occur together.

With relation to permeability enhancement, the size and isolation of pattern A fractures mean that any of the three fracture sets might predominate in a well. This lowers the predictability of the resulting directional permeability. For pattern B, however, as the figure shows, avenues for fluid communication will tend to parallel the trend of the host structure.

Relationship between Fracture Porosity and Permeability to Structure Curvature

In cases where pattern B is dominant, a set of simple expressions has been derived to relate fracture porosity and permeability to bedding thickness and structural

curvature (Murray 1968). These are given in Figure 5 (Part a reveals basic equations relating structural curvature and petrophysical character.

Figure 5

Part b is a graph showing influence of bed thickness on fracture permeability. Note that doubling this thickness increases permeability by a factor of ten.). The specific structure for which they were derived was an asymmetric anticline in the Williston Basin of North Dakota. A structural contour map on top of the productive Mississippian Bakken formation is shown in Figure 6 , and makes obvious the correlation between structural curvature and well productivity.

Figure 6

What various fracture patterns tell us in a more general sense is that the same body of rock is often subjected to several different states of stress during the folding process. In every case, therefore, the specific relationship between a fracture pattern and fold geometry must be carefully established, usually by statistical techniques.

Wellbore Breakouts and In Situ Stress

Within the past twenty years, regionally consistent patterns of borehole elongation have been observed in different provinces. More recently, the four-arm dipmeter caliper has been able to detect the long axis of such elongation. Images processed from ultrasonic borehole televiewer surveys have, meanwhile, revealed the actual shape and dimensions of the borehole wall. The televiewer is a tool that emits ultrasonic pulses from a rotating piezoelectric transducer at a rate of 600 per revolution. It produces an "unwrapped" image of the wellbore surface that has proved useful for identifying and studying natural fractures.

Wellbore breakout is the term used to describe the spalling of rock that appears to create elongation. To date, the data indicate that breakouts are relatively broad, flat curvilinear surfaces that enlarge the wellbore on opposite sides to produce a final

elliptical shape ( Figure 1 , Proposed cause of wellbore breakout).

Figure 1

The two most current and accepted interpretations of this phenomenon attribute breakout to (1) the intersection of the wellbore with natural fractures, and (2) compressive shear failure due to stress relief, such that the direction of elongation is parallel to the in situ minimum horizontal compressive stress ( Figure 1 ) (Zoback et al. 1985). A growing consensus based on recent analyses strongly favors the second interpretation.

The stress relief hypothesis is especially attractive, since it allows for relatively straightforward derivation of the basic in situ stress field. This can have obvious importance for explaining and predicting the orientation of hydraulically induced fractures in low permeability reservoirs.


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 trouble-causing termination of a coal seam. Thus, "fault" then

carried much of its vernacular sense; basically, some sort of a mistake had been made.

Geologists like Murchison and Lyell, however, were quick to realize that a fault was, in reality, a fracture along which 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 sink 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 above the fault and rest his foot on those 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. 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 fault could be found again if the miner continued his tunnel a short distance in the same direction and then sunk a shaft downward ( Figure 1 ). (In this area of Britain, such faults are far more common.) At times, the reverse was true, and thus the corresponding fault was designated as such. These terms remain in use.

Geologists have long known that faults can extend laterally for much greater distances than folds, and that they sometimes represent structures of far greater significance. This has been confirmed by the theory of plate tectonics, which has not only "discovered" and explained a new class of faults (transform faults), but has emphasized the importance of faulting over folding in general, especially with respect to deciphering the origin of tectonism.

Structural style often provides a broad context for understanding the patterns of faulting that may be expected in a region. At the same time, however, the explorationist must generate prospects, and this sometimes requires a good deal of knowledge about detailed lithologic and structural constraints.

Description and Basic Terminology

By modern definition, rocks are said to be faulted when they have suffered observable displacement along a plane or interval of rupture. Such rupture occurs mainly by shear. The fault plane can be relatively simple (part a of Figure 1 )

Figure 1

or it may consist of a large number of individual offset surfaces and thus be more accurately described as a fault zone (part b of Figure 1 ). Less frequently, rocks may become displaced by a form of shearing that causes loss of internal cohesion but not

actual rupture of lithological layers (part c of Figure 1 ). Geologists also use the term shear zone to label fault zones in which the individual planes of displacement are extremely closely spaced. The great majority of faults, however, more closely approximate planes of slip along which shearing has taken place as a result of movement.

The location, style, and geometry of faulting are due to the interaction between the applied stresses and the initial properties of the affected rocks. Inhomogeneities resulting from sedimentation or diagenesis, such as the abrupt facies changes at a reef margin or the draping of units over a barrier bar, can localize faulting. Faults are often genetically related to folding. A great number of possibilities exist, and it is often helpful to the explorationist to delimit those that appear relevant to a particular area, since this knowledge can be used to predict or explain the occurrence of faulting in other parts of an area.

Faults are most 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.

Net slip along a fault is measured by a vector that traces the displacement between originally adjacent points ( Figure 2 , Basic terminology for fault offset, showing strike slip (ss) and dip slip (ds) components of net slip (ns).

Figure 2

The fault itself is oblique slip; points P and P' were coincident before faulting.). It is

most often resolved into dip slip and strike slip components. Faults in which one or the other of these components is dominant are correspondingly named.

For non vertical dip slip faults, geologists find it useful to again divide the displacement into components, this time into vertical and horizontal contributions ( Figure 3 ).

Figure 3

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." These terms are somewhat archaic, and remain more frequently used in mining than in petroleum geology. Petroleum geologists, and structural geologists in general, more often 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. In addition, the normal stratigraphic separation, shown in Figure 3 , is often important to determine, particularly with reference to log data.

Part a of Figure 4 shows the four types of dip slip faults (Major types of dip slip faults.

Figure 4

H and F refer to hanging wall and footwall ). In most cases, it is assumed that simple shear has acted as the principal strain within the fault plane (part b of Figure 4 , Diagram illustrating the progressive simple shear within an ideal fault plane. Most faults have complex histories of offset involving intermittent, incremental, and pulse like periods of movement.).

It should be mentioned that 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 intercontinental basins are characterized by near-vertical, high-angle faults of small displacement that cannot be easily identified as normal or reverse. Furthermore, as mentioned, multiple episodes of tectonism often affect a single region. Such episodes may involve contrasting stress regimes, and displacement along early faults can be reversed. One of the more important occurrences of this phenomenon involves normal-fault-dominant provinces being subjected to compressive stresses as a result of changing plate boundary interactions. This creates "inverted" structures, which we shall look at.

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 unreformed sediments.

Most rupture along normal faults is by shear. Failure due to tension fracturing does occur, but only at shallow levels where confining pressures are very low. Tension fractures tend to be nearly vertical and are commonly removed by erosion. Only in areas where faulting has begun very recently (e.g., certain portions of Iceland) will such fractures be visible.

The term thrust fault deserves special discussion. A thrust has often been defined as a reverse fault dipping less than 45º However, as used by most geologists today, "thrust" is a genetic term, said to imply near-horizontal, tangential compression and a zone of movement dominated by simple shear strain (part b of Figure 4 ). Shortening is involved, and this can be as great as tens of kilometers or more for single faults.

Thrust faults can (and do) dip at any angle, and individual thrust planes often show complex, curving geometries. They may be essentially horizontal (along bedding planes) for kilometers, but curl up at their termini to become nearly vertical or even overturned. In the latter case, thrusts become apparent "normal" faults, geometrically speaking. Section 6.5 discusses these faults in detail.

In relation to both normal and thrust faults, we should become familiar with several other commonly used terms. These were originally derived by German geologists to describe the structural features that characterize the great valley of the Rhine, as well as those of the Bavarian and Austrian Alps. 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 (part a of Figure 5 ).

Figure 5

On the other hand, as shown in part b of Figure 5 , a cover of thrusted rocks can become locally eroded to the extent that a fenster or window to the underlying footwall lithologies is created, or an isolated remnant called a klippe is left. Either of these can range from tens of meters to hundreds of square kilometers in size.

Strike slip faults can be simply classified as left-lateral or right-lateral ( Figure 6 ) on the basis of the shear sense.

Figure 6

When facing the fault, the direction in which the opposite side appears to have moved gives this sense. The San Andreas ( Figure 7 ) is one of the best-known examples of an active right-lateral strike slip fault.

Figure 7

Others include the Alpine fault of New Zealand and the Atacama system of may be essentially horizontal (along bedding planes) for kilometers, but curl up at their termini to become nearly vertical or even overturned. In the latter case, thrusts become apparent "normal" faults, geometrically speaking. Section 6.5 discusses these faults in detail.

In relation to both normal and thrust faults, we should become familiar with several other commonly used terms. These were originally derived by German geologists to describe the structural features that characterize the great valley of the Rhine, as well as those of the Bavarian and Austrian Alps. 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 (part a of Figure 5 ). On the other hand, as shown in part b of Figure 5 , a cover of thrusted rocks can become locally eroded to the extent that a fenster or window to the underlying footwall lithologies is created, or an isolated remnant called a klippe is left. Either of these can range from tens of meters to hundreds of square kilometers in size.

Strike slip faults can be simply classified as left-lateral or right-lateral ( Figure 6 ) on the basis of the shear sense. When facing the fault, the direction in which the opposite side appears to have moved gives this sense. The San Andreas ( Figure 7 ) is one of the best-known examples of an active right-lateral strike slip fault. Others include the Alpine fault of New Zealand and the Atacama system of Chile (

Figure 8 and Figure 9 ,

Figure 8

Examples of major strike slip vaults in various parts of the world.

Figure 9

Dots mark the site of active volcanoes.). Left-lateral faults of such scale are well represented by the Great Glen fault of Scotland and the Philippine fault. Such large strike slip structures are often called wrench faults, in general, and transcurrent faults more specifically if they cut across regional structural trends. 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.

Fault Zone Material

Faults and fault zones are marked by crushed and sheared rock fragments, called fault breccia. In cases where this has been milled and broken down to a clay-rich powder, the term fault gouge is used. When relatively coarse and loosely cemented, breccia can serve as an excellent conduit for fluid migration. Gouge, on the other hand, is more likely to act as sealing material. The rocks immediately adjacent to a fault will also suffer a certain amount of shearing, which may fracture or weaken them. Breccia or gouge zones and their conductive or sealing capacity, therefore, sometimes show gradation into the surrounding country rock. This is particularly true for wide intervals of shearing associated with reverse faults of major displacement.

At deeper levels, the brittle crushing of rocks-- called cataclasis-- within the plane of actual movement will show a transition under increasing temperatures and pressures to more continuous recrystallization or actual flow. Geologists use the term mylonite

for the microbrecciated, often highly recrystallized and intensely sheared rock that forms within these zones of deformation. Mylonites will act to seal a fault from fluid migration.

Most gouge and breccia intervals are the result of several stages of movement; it appears rare that faults complete their total offset in one continuous episode. Most often, therefore, these intervals remain structurally weak. Where exposed at the surface, therefore, fault zones are usually topographical depressions.

On the other hand, it is also possible for faults to be tightly cemented by material that has precipitated out of mobilized solutions. Fault zone mineralization may produce lateral conductivity barriers composed of such highly resistant, impermeable material as quartz or siderite. Examples exist where differential erosion has removed the relatively weak country rock and left standing high-angle breccia zones that have the appearance of igneous dikes.

During the initial stages of movement, most faults probably provide conduits for the migration of fluids. Such fluids would essentially be drawn out of the immediately surrounding rock, but may, of course, travel substantial vertical distances within the fault zone itself. In general, the rupture, movement, and shifting of large an isotropic rock masses on either side of a fault often cause internal readjustment that includes the release and migration of mineral-rich solutions. Frictional heating as a result of shearing and pressure drop within the fault zone itself also contributes to the mobilization of such fluids and the subsequent precipitation of impermeable material (especially secondary clay, quartz, and calcite) from them (Ramsay 1967).

The degree to which such sealing occurs is of obvious importance to petroleum geology. The problem we face in predicting this is that a number of complex factors are involved: the type of fault; the detailed history of its displacement; the mineral, matrix, and pore fluid composition of the rocks immediately adjacent to it; and the relevant pressure/temperature conditions. These are only the most obvious. As a result, the sealing capacity of a fault most often varies, both along its vertical extent and along its strike. A few useful generalizations, however, can be made.

Normal faults, because they are the result of extension, are not generally so intensively brecciated or milled as reverse faults. Their planes of movement are relatively narrow, and high pressure effects, such as mylonitization, are usually lacking.

Reverse faults, by contrast, are likely to be wider for a given displacement and to involve more intense shearing for a given duration. It would appear that, in their earlier stages, thrust faults are highly conductive to fluids, but may just as often become sealed later on. Major thrust faults that cut at low angles through competent lithologies sometimes create wide zones of fracturing and brecciation that remain conductive. In addition, because they do not involve large offsets and are relatively shallow in occurrence, imbricates often do not mill the country rock to gouge, but instead produce zones of relatively porous breccia.

Thrusts "leaking" petroleum have been drilled in the western Canadian thrust belt, in the western over thrust belt of Wyoming and Utah, and in parts of the Rocky Mountain foreland of the western United States. In the Wyoming-Utah thrust belt, in fact, both major and secondary faults are known to have relatively high fluid conductivities. Productive reservoirs appear to exist where reservoir rocks are juxtaposed against Cretaceous source beds across fault planes. In the case of the

Rocky Mountain foreland, faults involve great thicknesses of basement crystalline rocks and are often relatively wide zones of coarse breccia.

Strike slip faults can show a complete spectrum from narrow gouge zones to mylonite intervals up to kilometers in width. The character of the material in the fault plane is, in a general way, related to the degree of compression involved and to the magnitude of total offset. Larger faults showing greater displacements are generally more mylonitized, or show other types of intense and laterally extensive deformation. The San Andreas, for example, is marked by wide zones of associated fracturing and brecciation, as well as mylonitization. Considerable variety is often seen along a single fault plane.

Minor Structures Associated with Faults

Like folds, faults are frequently large and complex enough in their deformation to generate a characteristic set of legible minor structures. These can often be of considerable help in the analysis of both individual and regional structures.

Slickensides

Perhaps the most common features associated with faulting are the polished and striated surfaces called slickensides that directly record the sliding movement of rock surfaces past each other. They commonly occur as parallel grooves, striations, stretched and flattened mineral grains (especially quartz and calcite), or as combinations of these that parallel the direction of relative motion when they formed. They have been observed along fault zones involving almost every type of rock.

Slickensides show a puzzling diversity in their specific characteristics that cannot be simply attributed to differences in rock type and fault movement. They often show characteristic steps ( Figure 1 ), and are sometimes penetrative over a narrow interval (Drawing of a slickenside sample in shale from southern New York Note the characteristic steps and ribbing of mineral grains.).

Figure 1

Within this zone, however, they can indicate more than one direction of movement. Furthermore, if the direction of fault movement changes, they will become overprinted.

As a phenomenon related to shearing, slickensides are not confined to fault zones, but are also associated with flexural slip folding, being most well-developed along contacts between thick, competent units. They can also occur along fractures called shear joints that have suffered small-scale offset.

Drag Folding

A second common minor structure associated with faults is known as drag folding. This is used to describe the warping of layers in the immediate vicinity of a fault zone. This term, despite its ubiquity, is often misleading, since it implies that faulting occurs first and causes folding by friction. In reality, as pointed out by Hobbs, Means, and Williams (1976) and Park (1983), ductile warping following rupture is far less likely than the reverse. ( Figure 2 , Schematic illustration of the "preferred" origin of fault-generated drag folding. Pre-faulting strain results in offset layers dipping up or down into the fault plane toward each other.)

Figure 2

Drag is described as "normal," then, if it accords with what would be expected if folding occurred first; "reverse drag," is in the opposite direction. Here, obviously, is another case where terminology can easily be confused: "normal" and "reverse" when used to describe drag folding have no connection to the sense of displacement along the relevant fault. In Figure 3 examples of normal drag for both normal and reverse faults are given.

Figure 3

Reverse drag is more commonly seen in association with listric normal faults; displacement tends to roll beds in the hanging wall over to form a gentle anticline. Such "rollover' " has proved extremely important to petroleum accumulation.

Transverse folds, sometimes of considerable size, as well as both normal and reverse faults are known to be associated with large strike slip faults such as the San Andreas. Figure 4

Figure 4

and Figure 5 relates the range of these parasitic structures to the mean strain ellipse of a major wrench fault.

Figure 5

Certain other secondary structures, also generated by folding, are found in association with faults, especially reverse faults. Shear and extension joints, for example, are frequently related to faulting. In addition, thrusting is capable of inducing cleavage in incompetent layers, both above and below the plane of movement (Wood 1974).

Basic Interpretation of Fault-Associated Structures

The specific suite of secondary structures associated with a particular fault is the result of a complex interaction among the following parameters: (1) the internal properties of the rocks involved; (2) the initial orientation of these rocks relative to the fault plane; (3) the temperature and pressure conditions of the surrounding subsurface environment; (4) the rate, magnitude, direction, and duration of the relevant stresses. Careful study of minor structures will, therefore, provide us with at least some amount of information about these factors. Obviously, such information can be useful to an understanding of how faulting may have affected a potential reservoir interval. Moreover, as in the case of folding, geometric relationships may also become clear from statistical analysis of secondary features. Fractures, micro-faults, minor folds, slickensides, fault cleavage, and gouge or breccia are readily discernible in cores, if these can be recovered. Fault gouge or breccia is often apparent during mud logging.

In the case of fractures and minor faults (fractures along which movement has taken place), their orientation with regard to the main displacement may or may not be simple and straightforward. Minor folds are most often caused by fault shear, and thus their axial planes should roughly parallel the fault plane itself. Cleavage that is generated by faulting often trends slightly oblique to the fault plane, however. If it cuts through several layers, the line that marks the intersection between cleavage and bedding planes should roughly parallel the line marking the intersection of the fault with bedding

Principal Stress Directions in the Development of Faulting and Fracturing

Figure 1 displays the idealized orientation of the principal stress axes during reverse, normal, and strike slip faulting.

Figure 1

Because faulting and fracturing both represent the brittle rupture of rocks, they are often discussed in relation to similar stress systems.

Figure 2 shows the generalized orientation of actual fractures formed in an experimental triaxial test, in which a block of Solenhofen limestone was shortened by

1 % at room temperature.

Figure 2

The two categories of fractures that geologists 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 said, rocks are weakest in shear.

Two mutually orthogonal types of extensional ruptures are shown in Figure 2 These have been related to the stresses generated during loading and, 1 and 3 unloading, and, in each case, they develop normal to 3. Notice that during unloading exchange the orientations they had during loading. What this tells us is that faults can conceivably develop at very low angles to 1 , 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 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 stress axes alone.

At smaller and smaller levels of scale, many faults reveal increasingly complex components of displacement, usually involving shear couples. Figure 3 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 3

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. As with folds, the summation-of-local-strains method is often the most accurate approach to analyzing complicated fault systems.

In Figure 3 , three sets of fractures are evident: two seem to be conjugate shear fractures; the third, a bedding plane extensional fracture. If our scale were on the

order of miles instead of meters, there would most likely be smaller, third-order structures associated with the shear fractures. Thus, we would build our portrait of the overall structure in steps, beginning with what we could observe and measure most easily.

Figure 4

Figure 4

and Figure 5

Figure 5

give examples of regional strike slip and normal faulting ( Figure 4 shows the resolution of maximum and minimum principal stresses for the San Andreas fault system, California. Figure 5 is a generalized cross section through a region of extensional (rift) deformation showing the resolution of maximum and minimum principal stresses.). Reverse faulting on a comparable scale occurs most often in long and narrow thrust fault provinces called fore/and thrust and fold belts. ( Figure 6 , Generalized thrust best structure showing variety of faults and folds and the main types of associated petroleum traps.

Figure 6

Cross section covers horizontal distance of about 100-150 km and is based on detailed studies of the Canadian Rocky Monitions.) Significant hydrocarbon entrapment occurs in such provinces throughout the world. Given this and the highly complex nature of the structures involved, we should look in more detail at the nature of thrusting.

Thrust Faults

For over a century, the existence of low-angle fault planes, along which rocks have sometimes been transported for up to 100 km, has baffled and intrigued structural geologists. Mechanisms proposed for such movement have included rigid push, large-scale gravity sliding, and, more recently, convergence between plates (at subduction zones) and within plates, as well as collision between continental portions of plates (see Figure 1 , Generalized map and cross section showing continental breakup along the Red Sea rift and collision in the Zagros region of southeastern Iran.

Figure 1

Numbers indicate total estimated separation – in km – between Africa and Arabia).

The debate continues for particular provinces, although the last three mechanisms are now generally regarded as most important.

The Concept of Relative Shortening

Given the structures of foreland thrust and fold belts above a regional plane of detachment ( Figure 2 , Generalized thrust belt structure showing variety of faults and folds and the main types of associated petroleum traps.

Figure 2

Cross section covers horizontal distance of about 100-150 km and is based on detailed studies of the Canadian Rocky Mountains.), it does not seem possible to distinguish between underthrusting- in which the lower, relatively undeformed "block" is active-and overthrusting-in which it is the upper block, in which deformation is concentrated, that actually moves. Relative shortening is what geologists measure in thrust terrains.

In foreland belts, this has been determined to range up to 150 km (Price and Mount-joy 1970), and is probably greater in certain mountain systems, such as the Himalayas. To some degree, these provinces pose the culminative challenge to structural geologic analysis, involving, as they do, the entire spectrum of major and minor structures seen in nonmetamorphosed sedimentary rocks. Nearly all styles of folding and faulting are present and intimately related. This has both created structural traps for giant petroleum accumulations as well as destroyed significant hydrocarbon potential by breaking up and exposing source and reservoir rocks. To understand how the basic features of thrusting are described and analyzed by geologists, we should become familiar with a few more commonly used terms.

Thrust Fault Geometry and the Influence of Lithology

Rocks that have been transported from their original location ("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." Most thrusts define a listric plane that flattens with depth. Several or more subsidiary thrusts commonly occur within a single allochthon, either as splays off the "sole" fault 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 ( Figure 3 , Simplified map and cross section from same general areas as Figure 2, showing complex thrust-fold deformation in the southern Canadian Rockies.

Figure 3

Note the apparent reversals in the direction of thrusting due to refraction of fault planes through lithologic sequence of variable ductility. Other features on the map include a Klippe (K) and an anticline that becomes a thrust along strike (A)).

The actual zones of movement may be mylonitic at depth, or, if the thrust passes through a ductile unit such as a thick shale, very narrow, unbrecciated, and essentially parallel to bedding for distances that can range up to tens of kilometers. 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 and Figure 4 ,

Figure 4

Figure 5

Figure 5

and Figure 6 summarize this influence (Figure 4: Diagrams illustrating present-day geometry along a major thrust fault in the Canadian foreland and a detailed reconstruction of its upward path through various stratigraphic units.

Figure 6

Figure 5: Diagram using the concept of fault preference to show how thrust development is often directly a function of stratigraphy. Lithologies shown are those of Figure 4. Fault preference is defined as the length of a thrust within a unit divided by the thickness of that unit, multiplied by an arbitrary constant. Note here the strong preference at greater depth for major thrusts to exist in the Kootenay and Fernie shales. Figure 6: Ramp model for thrust development. Note that it is the lower plate that is shown to move):

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)

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

the age of major faults is younger in the direction of thrusting

major thrust faults do not overlap appreciably

The Evolution of Thrusting: Ramping, Decollement, and Imbrication

In most fore-land belts, therefore, the overall evolution 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 stair-step evolution of thrusts, often called ramping ( Figure 6 ) A relatively flat portion of a thrust plane, particularly where it remains parallel to bedding within a single lithology, is referred to as a decollement, a term we have used previously to describe the basal detachment that sometimes occurs in parallel folds.

The progressive stacking of thrust faults may develop in "piggyback" fashion-where younger thrusts form in the foot-wall-or alternatively in "overstep" fashion, where thrusts become younger toward the root zone. As mentioned, piggybacking appears to prevail 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 concentration: (1) near the toe of a major thrust, and (2) 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 listric faults, which sole out into a major thrust plane. Continued movement along this plane after imbricates have formed will rotate them so that 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 (see Figure 2 ).

Thin-versus Thick-Skinned Tectonics

The concept of basal decollement is the fundamental structural principle in the hypothesis known as thin-skinned tectonics; this is often contrasted with the thick-skinned hypothesis, which postulates no sole fault and, therefore, direct involvement of basement in each major thrust ( Figure 7 , lower section, Cross sections illustrating the thin-skinned (upper) and thick-skinned (lower) hypotheses for regional thrust faulting in the Appalachians).

Figure 7

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 geologists now favor the thin-skinned hypothesis for at least the more medial and distal portions of thrust belts. However, it is well known that toward the metamorphic core of many such belts, basement rocks are heavily involved in thrusting. Such involvement, as in the case of the Himalayas, can be very extensive and may, in fact, control the overall style and evolution of thrusting.

Yet some recent studies based on deep-reflection seismic profiles (Cook 1982; Cook et al. 1979) have strongly favored the thin-skinned hypothesis for basement thrusting as well. Thus, the debate has expanded to focus on two major questions: (1) whether thick-skinned faulting occurs at all in foreland belts; and (2) 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.

Finally, as we have mentioned, the development of foreland thrust and fold belts seems to occur in pulses. Each episode intensifies the deformation in existing structures and also generates new features that modify these structures. This superposition of structures can create a number of seemingly anomalous age and

thickness relationships. Figure 8 displays several of these.

Figure 8

Tear Faults

During its movement, a large thrust sheet or reverse-faulted block may tear along near-vertical planes oriented transverse to the principal tectonic transport direction ( Figure 9 ,

Figure 9

Orientation of major tear faults (numbered dashed lines) in the Jura mountains, French-Swiss border and Figure 10 , Tear faults along the northeast corner of the Beartooth Mountains, Montana). Normally, these either form in this direction or at angles to it that invite explanation of them as conjugate shear faults.

Various genetic schemes have been proposed for tear faults. They have been mapped as striking at both low and high angles to regional , and thus appear to be generated by compression or extension within the allochthon. These stresses can be explained as the result of differential movement. Due to a variety of factors, such as stratigraphic changes along strike, large thrust sheets apparently advance at different rates over portions of their length (Elliot 1976). In addition, the direction of transport can also vary along strike (see Figure 10 ); thrust terminations often show bow-shaped patterns that indicate a type of radial displacement (Dahlstrom 1970; Davis 1984; Lowell 1985).

Figure 10

Transform Faults

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 such 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.

The currently accepted interpretation of transform faults was first put forth by Wilson (1965), who sought to reconcile the offset of ocean ridge segments and magnetic anomalies with earthquake first-motion data. Such data indicated that actual movement along the faults was opposite to that indicated by the apparent physical displacement of ridge segments. Wilson's analysis showed that this displacement represents an original "frozen" geometry of continental separation. Actual fault motion due to sea-floor spreading, however, has the opposite sense ( Figure 1 ,

Figure 1

Schematic illustration of transform fault development and Figure 2 , Block diagram showing relationship between rifting and transform faulting.

Figure 2

Note that offset of rift segments is opposite to the direction of actual displacement across the transform fault (shown by arrows). Crustal attenuation is also illustrated.). Continued spreading, then, will not affect the offset of ridge segments.

Such transform faults have come to be known as ridge-ridge transforms, and represent one of several possible types of plate-boundary connections. Figure 3 shows two other types of transforms and the resulting displacement across them.

Figure 3

In Figure 4 , we see examples of a ridge-ridge and ridge-trench transform (San Andreas and Fair-weather fault systems, respectively).

Figure 4

Note that the northern corner of the Cocos plate must also be marked by a ridge-trench transform.

It is speculated that ridge-ridge transforms act to decrease the dynamic resistance to spreading (Bally and Oldow 1983). This resistance is at least partially amplified by differential rates of spreading along a ridge axis. The faults are said to provide avenues of low shear resistance for separation, and to segment the diverging plate margins into orthogonal steps that reduce the effective surface area. They trend at right angles to ridge segments and therefore trace the actual direction of plate motion. With respect to subduction and collision boundaries, transforms appear to develop in response to the differential velocities and directions of converging plates. As such, they frequently represent a type of metastable condition that will continue until major plate reorganization (Bally and Oldow 1983).

Where transform faults cut continental crust, the term wrench fault is used. The San Andreas fault is a well-known example that connects the Juan de Fuca Ridge off the U.S. Pacific Northwest with the Baja spreading ridge in the Gulf of California. This type of transform is of some importance to petroleum exploration. Associated folding and normal faulting have created structural traps for large hydrocarbon accumulations, such as those found in the Bakersfield area of southern California (Harding 1976).

Basic Influence of Faulting on Logs

Figure 1 ,

Figure 1

Figure 2 ,

Figure 2

and Figure 3 show the predicted effect of normal and reverse faulting on log curves, along with two field examples.

Figure 3

Figure 3 shows a cutout of section from a well in Alaska's giant Prudhoe Bay field. Figure 2 is from a well in western Iran and shows, on the other hand, the repetition of section in gamma ray, resistivity, and dipmeter logs caused by relatively low-angle thrusting. Note how the dip log responds to drag immediately below the fault. An anticline in the upper limb is clearly indicated by the upward-decreasing dip pattern, while the fault plane itself is marked by a zone of chaotic dips.

Basic Concepts

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 measurement complex is the variety in the components of force (with their changing magnitudes and directions) that can 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.

For 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 types: (1) body forces, which act at

every point within the crust; and (2) 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. On this grain, two types of body forces act at all times: gravity and inertia. The force due to gravity is called the lithostatic pressure, which results from the simple weight of overburden transferred by grain-to-grain contact. If our particular grain becomes involved in diastrophism, two types of surface forces will also exert their influence on it: pressure forces, which act perpendicular to the surface of the clast, and viscous forces, which act parallel to these same surfaces (Elliott 1976). (Note that viscous, as used here, does not imply anything about the nature or behavior of the material composing the clastic particle.)

Figure 1 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 surface forces that are primarily responsible for the creation of geologic structure and with which the structural geologist is concerned.

A rock undergoes deformation when stress causes the displacement of particles within it. Such stress can result from body or surface forces, or both. Deformation in nature 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.

Geologists 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 sizeable 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. (We say "approximates," since there is nearly always some component, however small, of lateral stress in the crust.)

The principal deformational effect of this body force is the compaction of sedimentary grains. However, simple overburden can also create abnormally high pore fluid pressures that, in turn, may lead to slumping and "growth" faulting.

The Three Principal Stresses

As a measure of force-per-area, stress is a vector quantity and thus may be expressed as the sum of various components. ( Figure 1 , Resolution of unidirectional force (F) acting on a cube face into the basic normal (N) and shear (S) components of stress.

Figure 1

Also shown is the idealized physical effect of each component.) If we trade our clastic grain for a hypothetical cube of uniform composition and subject this cube to a progressive deformation, we see that the total force (F) can be resolved into contributions that act perpendicular to the faces of the cube, and along (parallel to) them. ( Figure 1 and Figure 2 , Resolution of normal and shear stresses on a cube undergoing simple progressive deformation.

Figure 2

The direction and magnitude of force (F) remain constant. As deformation progresses, the relative magnitude of the shear stress decreases, while that of the normal stress increases.)

On every face of our hypothetical cube, then, there will usually exist both normal stresses (the pressure-force- per-unit area) and shear stresses (the viscous-force-per-unit area). Normal stresses can be either compressional or tensional and tend to compact or separate particles within a body. The effect of shear stresses, on the other hand, is to move particles past one another ( Figure 1 ). During deformation, the relative magnitudes of these two stress types will change, such that one type may decrease as the other increases ( Figures 2 ).

Materials 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 thus only the components of normal stress exist. These three planes are known as principal planes of stress, and the axes of their intersection are thus the principal axes of stress. These axes, then, 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 (see Figure 3 ,

Figure 3

Figure 4 and Figure 5 ), i.e., compression ("squeezing") or tension ("pulling-apart").

Figure 4

(Generalized orientation of the principal axes of stress for three basic geologic structures. In Figure 3 and Figure 4, horizontal compression can be thought of as the actual applied force, while Figure 5 shows the effect of horizontal extension.

Figure 5

)

In geology, principal stress is always spoken of in terms of compression, which is taken as positive (for materials and engineering science, the opposite is true, i.e., tension is positive, compression negative). The three principal axes, or stress directions, are correspondingly written as 1 (maximum principal stress), 2 (intermediate principal stress), and 3 (least or minimum, principal stress). For the purposes of structural geology, 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 1 = 3 and is nonzero, while 2 is zero. This is a special case of biaxial stress.

Within the earth's crust, the most common stress situation is triaxial, with 1 > 2> 3 >0.

In terms of actual physical effects, 1 and 3 can be thought of as representing the relative compression and tension that act to deform a rock body by compacting particles into each other and stretching them away from each other. We say "relative" because, very often, these stresses are defined only in relation to each other. For example, a strong extensional event in the crust will have a 1 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 , Figure 4 and Figure 5 which place our hypothetical cube in evolving fold and fault structures. Again, 1 represents the direction of maximum effective compression; 3, 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 they, and 2, are normal to the plane of the paper. In physical terms, 2 is parallel to the surface of the earth. It 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 our cube, the fact that 1, 2 and 3 differ in magnitude and direction means that shear stress is generated within the cube. In fact, the quantity 1- 3 (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 inclined at 45° to 1 and 3 that include 2. ( Figure 6 , Diagram showing planes of maximum shear stress in an idealized triaxial stress system that is applied to a homogeneous block of material.

Figure 6

Note that these planes are inclined at 45 between any two principal stress axes.) 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.

Despite the extensive and detailed literature available on stress analysis in various rock types, little is known about stress fields that govern deformation in progress. Though techniques exist to determine in situ stresses, these only give us information about the final structure and the stresses either stored within it or associated with its specific setting within a larger tectonic environment. Such techniques do not directly tell us about the evolution of a particular structure.

Figure 7

Figure 7

and Figure 8 show two computer-generated interpretations of normal and shear stresses in a folding layer that has already suffered approximately 630/c total shortening.

Figure 8

The short lines in Figure 7 are drawn perpendicular to 1 and therefore closely approximate 3. We might just quickly note three important aspects to Figures 7 and 8: (1) how 3 changes direction over different parts of the folding layer; (2) how this compares to the changing directions of shearing stress shown in Figure 8 ; and (3) how the sense of these shearing stresses reverses across the folding layer. All of these reveal significant elements in the folding process.

Basic Concepts

Strain is said to exist when particles within a body have undergone displacement: specifically, 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). As implied by our previous discussion, strain can result from both body and surface forces. It may also, however, be caused by changes in the temperature and chemical composition of a rock body.

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 and Figure 2 (Computer-generated stress field for a hypothetical fold in a less viscous matrix.

Figure 1

The short lines in Figure 1 represent 3 axes and in Figure 2,

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 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, geologists often introduce helpful simplifications into their 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.

Types of Strain

Much of the terminology derived for understanding stress has also been applied to strain. 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 (part a of Figure 1 ).

Figure 1

Thus, for example, a cube becomes a rhombohedron, while a sphere inscribed within it becomes an ellipsoid.

This should also help make clear the basic concept of inhomogeneous strain (part b of Figure 1 ), 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.

Such fabrics often provide invaluable clues to both small- and large-scale structural patterns. Natural fracturing, of course, is of particular importance to petroleum geology, and knowledge of the stress-strain relationship associated with it can be very useful. As we will see, fracture patterns very often have a direct causal relation to major structures, such as faults and folds. The summation-of-local-strains method, therefore, will usually reveal this and permit the geologist to predict patterns in adjacent, undrilled areas. Several examples of this are given later on, when we look at fracturing in detail.

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 2 .

Figure 2

(Several basic types of homogeneous strain imposed on a cube of ideally uniform composition. In each case, the inscribed circle and ellipse represent cross sections through the strain ellipsoid before and after deformation. The types of strain shown are: a. uniform extension; b. uniform flattening; and c. plane strain. Note that no strain occurs in the intermediate direction.)

To understand how geologists treat natural deformation, however, it is also necessary that we look at the two basic types of shear strain. ( Figure 3 , Hypothetical cross section and diagram to illustrate domains of pure and simple shear in a series of folds that show progressively greater total strain.

Figure 3

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 often referred to as an irrotational deformation." It results from uniform extension in one direction and contraction perpendicular to it ( Figure 4 , Particle paths in simple and pure shear. Note rotation involved in pure shear). Strain that approximates pure shear is seen in many folds.

Figure 4

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 4 ).

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.

Strain Ellipsoid

Let us refer again to the sphere inscribed within our cube that undergoes homogeneous strain. Giving this sphere axes 1, 2 and 3, we find that these remain orthogonal through the homogeneous deformation and can be used to describe the long, medium, and short dimensions of the resulting ellipsoid. ( Figure 1 , Several basic types of homogeneous strain imposed on a cube of ideally uniform composition.

Figure 1

In each case, the inscribed circle and ellipse represent cross sections through the strain ellipsoid before and after deformation. The types of strain shown are: (a) uniform extension; (b) uniform flattening; and (c) plane strain. Note that no strain occurs in the intermediate direction.)

Together, these are called the principal strain axes of the strain ellipsoid. Just as stress can be said to exist at every point within a body, so is there a corresponding strain ellipsoid for these points, once deformation has taken place. Thus, comparison between the "before" and "after" shape and axes of the sphere inscribed within our cube provides us with a measure of the amount and type of strain.

It is standard practice for geologists to derive the principal axes of stress by superimposing them on the strain ellipsoid. ( Figure 2 , Note that this superposition assumes homogeneous strain, i.e., maximum shortening occurs in the direction of maximum principal stress, and maximum extension in the direction of minimum principal stress).

Figure 2

This is an optimistic simplification (as we have seen, the principal axes of stress and strain coincide only under conditions of homogeneity), but can be very useful. It has, for example, offered considerable insight into basic mechanisms and patterns of deformation-particularly faulting and fracturing-on many scales. This will become increasingly clear in the following two sections on folding and faulting.

Because of their frequent use of structural cross sections, geologists have also found it advantageous to make use of a strain ellipse-essentially a cross section through the strain ellipsoid along the 1 3 plane (i.e., the one that involves 1 and 3). The justification for this is, again, dependent on the assumption of homogeneous strain. Because of the regional nature of most tectonism, and the layered nature of lithologic sequences, many geologic examples of strain can be considered to approximate plane strain (see part c of Figure 1 ). In this type of deformation, the intermediate axis remains the diameter of the "original" sphere (the 2 axis-- parallel to 2-- in part c of Figure 1 ), while shortening and stretching occur along the other two axes.

Thus, two dimensions are sufficient to describe the strain at a particular point. If we are ready to accept the assumption of homogeneous strain, the strain ellipse becomes one of our principal indicators for the summation- of-local-strains method. ( Figure 3 Hypothetical cross section and diagram to illustrate domains of pure and

simple shear in a series of folds that show progressively greater total strain.

Figure 3

The shape of the strain ellipse can be the same for either type of shear and cannot be used to derive detailed strain history.)

As shown by Figure 4 , (Example of how the strain ellipse – here constructed from deformed oolites – can be used as a descriptive guide to deformation intensity and orientation.

Figure 4

Shown is a cross section through the south Mountain fold of western Maryland, U.S ) the strain ellipse can be an important guide to the general degree and style of deformation.

Some natural materials, such as ooids, spherulites, pebbles, certain fossils, and reduction spots in shales, can be used as qualitative ellipses or, in some cases, ellipsoids. However, because volume changes frequently occur during deformation (especially in carbonates) any quantitative determinations of strain based on such materials must be used with caution.

In principle, any object whose initial shape is known can act as a strain indicator. Such an indicator can be important to the subsurface explorationist, since it may be the only direct evidence available for how much strain has affected the fabric-and thus porosity and permeability-of a lithologic section. In most cases, the degree of tectonic influence on grain texture is fairly apparent from petrographic study. Strain indicators are primarily useful where this may not be clear and where special circumstances warrant mathematical determinations of strain. Specific techniques for measuring finite strain from oolites and spherulites are given by Ramsay (1967) and Ramsay and Huber (1985).

Nearly all deformation in nature is in-homogeneous. Not only do originally planar surfaces become complexly curved, but volume changes that involve both loss and addition of material frequently take place. Because of their pronounced heterogeneity in composition, thickness, and thus strength, rocks do not behave passively during deformation, but adjust in complex ways. Some units become strain-hardened and are able to withstand and transfer greater and greater amounts of stress as deformation progresses; other lithologies, in contrast, are fated to absorb stress by flowage, recrystallization, and the development of secondary fabrics such as cleavage.

Again, despite the dominance of inhomogeneity in nature, both local and regional deformational history can be reconstructed by assuming near-homogeneous strain domains. On a large scale, this often establishes the regional nature of stress and

strain. Geologists often estimate a regional strain ellipse based on the orientation of major structural trends. This is called the mean strain ellipse and is often useful in explaining such trends in terms of plate interactions

Rock Strength

A substantial literature exists with regard to experimental "rock crushing" in the laboratory, the intent of which is to simulate and analyze the process and effects of deformation. Studies are usually performed on cylindrical, corelike 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 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:

ElasticDuring 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).

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), and beyond it material is said to behave plastically; any increase in stress brings a corresponding increase in deformation ( Figure 1 , segment B).

Rupture With continued increase of stress, the body will eventually fracture (rupture) ( Figure 1 , point C).

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 (Stress-strain relationships for several ideal materials: (a) Hookean (elastic) body; (b) St.

Figure 2

Venant (plastic) body; (c) Newtonian (viscous) body ). Each style of behavior is denoted by the name of a well-known mathematician. Each 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 part c of Figure 2 , this type of body is represented as a porous piston pulled through a fluid-filled 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.

These three types of ideal bodies, then, 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 3 , which describe two generalized stress-strain relationships for known materials.

Figure 3

The principal difference between ideality ( Figure 2 ) and reality ( Figure 3 ) lies in the behavior known as strain hardening.

Such hardening is often the result of complex readjustment-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, 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 3 , 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 3 , 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 material ruptures 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.

In every situation, whether a rock responds in a brittle or ductile manner depends on several major controls: composition, effective confining pressure, temperature, and strain rate. It is, in fact, relatively meaningless to speak of the true "strength" of a rock without reference to these parameters. 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 ( Figure 1 and Figure 2

Figure 2

, Charts comparing the rates of common sedimentary and tectonic processes in various parts of the world), there is some transitional depth range over which the response of a particular rock type will grade from dominantly brittle to ductile.

Figure 1

Figure 3 is a schematic representation of the spectrum from brittle to ductile behavior in limestone, as determined by laboratory tests.

Figure 3

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.

It should not be assumed that brittleness guarantees faulting, or that ductility inevitably leads to folding. Most lithologic sequences are highly anisotropic, and, as a result, respond to tectonic stress in a manner that combines both styles of behavior. For example, the very shallow Shaur anticline amply demonstrates relatively brittle bending, which involves rigid body rotation and translation at shallow levels in response to high strain rates. At the same time, the extensive fracture system that characterizes this and many of the Zagros folds-and, in fact, is responsible for the very high rates of production (up to 80,000 bbl per day per well) in the Iranian Asmari oil fields-indicates simultaneous brittle rupture.

Figure 4 ,

Figure 4

Figure 5 , and Figure 6 indicate how laboratory analysis has shown deformation to vary as a function of pressure,

Figure 5

temperature, and strain rate.

Figure 6

The other factor that determines a rock's response to stress is, in one sense, the most obvious-- its composition. Figure 7 is a diagram derived by Handin et al.

Figure 7

(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 intracrystal line 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 geologists make use of the terms compressive strength, tensile strength, and shear strength, three major points should be made in relation to these. First, a material that behaves in a brittle manner can fail only under tension or shear. Second, almost all materials are weaker under tension than

compression ( Table 1 , below). And third, it is very often the shearing strength of materials that determines when and how they will fail.

Average crushing strength

Tensile strength

Shearing strength

Sandstone 740 10-30 100-200Limestone 960 30-60 150-250Quartzite 2020 30-90 100-300Granite 1480 30-50 150-300Basalt 2500 ------- 50-150

Table 1 . Measured strength of various rock types at standard temperature and pressure. Note the relative similarity in tensile and shear strength among materials with widely differing crushing strength. Note also the data for basalt. (Courtesy AGI)

The Mechanical Properties Log as an Estimate of Formation Strength

The mechanical strength of a particular formation can be estimated from a mechanical properties log, such as that in Figure 1 .

Figure 1

It can be very important to determine the strength in reservoir sandstones in order to prevent sand production problems. Both experimental and field test data indicate that a good correlation exists between rock strength, rock density, and certain dynamic elastic constants derived from compressional and shear velocities as recorded on sonic logs. Intrinsic strength is dependent on several major factors, including grain texture, cement, and the effective stress, which is defined as the overburden pressure minus the pore pressure.

In the example shown, G represents the shear modules, defined as the applied stress divided by the resulting shear strain; Cb is the bulk compressibility, defined as dilatation divided by the hydrostatic pressure; andz is the total intergranular space, derived from the interpretation of sonic logs. As explained by Tixier, Loveless, and Anderson (1973), the quantity G/Cb is the most sensitive indicator of formation strength. Note the corresponding changes in all log curves from the weak sand to the relatively tight lime unit below.

The Concept of Rock Competence

In addition to the concepts of brittleness and ductility, structural geologists frequently describe the relative strength of rock material in terms of competence.

This term is used to indicate contrasting behavior, or "ductility contrast," within a layered sequence of different rock units. Competent units are those that deform in a more brittle manner relative to incompetent units, which show more ductile behavior. A simple example of ductility contrast would be an alternation of thick sandstone and shale layers. During compression, folding of the competent sand units would control the deformation in the weaker incompetent shales, which tend to flow and accommodate the evolving fold geometry.

"Competent" and "incompetent," therefore, are strictly qualitative and, more importantly, relative terms. Looking at Figure 1 we can see that while limestone and dolomite share similar competencies in relation to quartzite at shallow levels, they differ dramatically at depth.

Figure 1

Below 4 km, limestone becomes highly incompetent relative to both quartzite and dolomite.

Conjugate Fracturing and Shear Couples

During experimental rupture tests (part (a) of Figure 1 ), rock often fails in the form of conjugate fractures.

Figure 1

(Part (a): Shear fractures typical of failure in dry limestone and marble: 1—Marble, brittle failure at 25 C, 35 bars, and 1% strain; II—Marble, "transitional" failure at 25 C, 280 bars, 20% strain; III—Solenhofen limestone, near-ductile failure at 25 C, 1000 bars, 11.2% strain; IV—Solenhofen limestone, ductile behavior followed by rupture at 150 C, 6500 bars, 9.1% strain. Part (b):Diagram illustrating angles of rupture in relation to maximum and minimum axes of stress.).

These represent full development of the discontinuities depicted in Figure 2 .

Figure 2

(Schematic diagram showing the spectrum from brittle fracture to ductile flow during typical compression and extension experiments on rock material. Typical strains before rupture as given, as are stress strain curves for uniaxial compression and extension. Ruled portions on the latter indicate the variation common for each case.).

The angle fractures make with respect to 1 and 2 is almost never 45° (see part b of Figure 1 and Table 1, below). Rock is generally a heterogeneous material and a great variety of internal physical properties determine the actual orientation of fractures. Generally, however, fractures occur at low angles (roughly 15-30 degrees) toand complementary high angles (60-75 degrees) to 3. Part a of Figure 1 portrays several examples of faulting created during testing. Table 1, below, gives the predicted angles of fracturing for common sedimentary and igneous lithologies. Comparing these data with those presented in Table 2, below, it would seem that a general correspondence exists between fracture orientation and shearing strength. This, of course, is not surprising.

Movement along conjugate fractures takes place by shearing, and the opposing motion across them is said to represent a shear couple. It is important to remember the basic orientation of 1, 2 and 3 in relation to shearing, since shear fracture is one of the most common types of strain observed in rocks. The effects of shearing are well documented on all scales, from displacement within calcite crystal lattices to the motion between lithospheric plates. Figure 3 and Figure 4 (San Andreas fault

system,

Figure 3

California) offer two examples of regions where the resolution of principal stress directions is fairly straightforward.

Figure 4

Geologists have found it useful to delimit shear couples in a wide variety of tectonic environments. In particularly complex regions, recognition of direction and attitude of the principal stress axes has proved its worth as both an explanatory and predictive tool.

Table 1. Predicted angles of rupture in relation to principal axes of compression and tension. Mean shear strength values are from Table 2. (Courtesy AGI)

Rock Type Main shear strength kg/cm2Shale (calcareous) 13° 77°Siltstone 20° 70°Sandstone 21° 69° 150Limestone 200 fine-grained 16° 74° oolitic 23° 67°Granite 17° 73° 275Basalt 21° 69° 100

Table 2. Measured strength of various rock types at standard temperature and pressure. Note the relative similarity in tensile and shear strength among materials with widely differing crushing strength. Note also the data for basalt. (Courtesy AGI)

Average crushing strength kg/cm2

Tensile strength

Shearing strength

Sandstone 740 10-30 100-200Limestone 960 30-60 150-250Quartzite 2020 30-90 100-300Granite 1480 30-50 150-300Basalt 2500 ------- 50-150

Definitions and Concepts

Many terms are used to describe the various parts of a trap. The anticlinal trap, the simplest type, will be used as our reference ( Figure 1 , Nomenclature of a trap using a simple anticline as an example).

Figure 1

The highest point of the trap is the crest or culmination. The lowest point is the spill point. A trap may or may not be full to the spill point. The horizontal plane through the spill point is called the spill plane. The vertical distance from the high point at the crest to the low point at the spill point is the closure. The productive reservoir is the pay. Its gross vertical interval is known as the gross pay. This can vary from only one or two meters in Texas to several hundred in the North Sea and Middle East.

Not all of the gross pay of a reservoir may be productive. For example, shale stringers within a reservoir unit contribute to gross pay but not to net pay ( Figure 2 , Facies change in an anticlinal trap, illustrating the difference between net pay and gross pay). Net pay refers only to the possibly productive reservoir.

Figure 2

A trap may contain oil, gas or a combination of the two. The oil-water contact, OWC, is the deepest level of producible oil within an individual reservoir ( Figure 3a , Fluid contacts within a reservoir in an oil-water system).

Figure 3a

It marks the interface between predominately oil-saturated rocks and water-saturated rocks. Similarly, either the gas-water contact, GWC ( Figure 3b , Fluid contacts within a reservoir in a gas-water system), or the gas-oil contact, GOC ( Figure 3c , Fluid contacts within a reservoir in a gas-oil-water system) is the lower level of the producible gas. The GWC or GOC marks the interface between predominately gas-saturated rocks and either water-saturated rocks, or oil-saturated rocks, as the case may be.

Before the reserves of the field can be calculated, it is essential that these surfaces be accurately evaluated. Their establishment is one of the main objectives of well-logging and testing.

Oil and gas may occur together in the same trap as separate liquid and gaseous phases. In this case, gas overlies oil because of its lower density. Source rock chemistry and level of maturation, as well as the pressure and temperature of the reservoir itself, are important in determining whether a trap contains oil, gas or both.

In some oil fields (e.g. Sarir field in Libya), a mat of heavy tar is present at the oil-water contact. Degradation of the oil by bottom waters moving beneath the oil-water contact may cause this tar to form. Tar mats cause considerable production problems because they prevent water from moving upwards and from displacing the produced oil.

Boundaries between oil, gas and water may be sharp ( Figure 4a ,

Figure 4a

Transitional nature of fluid contacts within a reservoir-- sharp contact) or gradational ( Figure 4b , Transitional nature of fluid contacts within a reservoir-- gradational contact). An abrupt fluid contact usually indicates a permeable reservoir. Gradational contacts usually indicate low permeability reservoirs with high capillary pressure.

Directly beneath the hydrocarbons is the zone of bottom water ( Figure 5 , Nomenclature of underlying reservoir waters). The zone of edge water is adjacent to the reservoir.

Figure 5

Fluid contacts in a trap are almost always planar but are by no means always horizontal. Should a tilted fluid contact be present, its early recognition is essential for correct evaluation of reserves, and for the establishment of efficient production procedures.

One of the most common ways in which a tilted fluid contact may occur is through hydrodynamic flow of bottom waters ( Figure 6 , Tilted fluid contact caused by hydrodynamic flow).

Figure 6

There may be one or more separate hydrocarbon pools, each with its own fluid contact, within the geographic limits of an oil or gas field ( Figure 7 , Multiple pools within an oil and gas field). Each individual pool may contain one or more pay zones.

Figure 7

Remember, the ratio between gross pay and net effective pay is important and is generally mapped from well data as the field is developed.

Classification

There are many different types of hydrocarbon traps. Several classification schemes have been proposed (Clapp, 1910, 1917; Lovely, 1943; and Hobson and Tiratsoo, 1975). Basically, traps can be classified into four major types: structural, stratigraphic, hydrodynamic and combination ( Table 1., below ).

Table 1. Classification of Hydrocarbon Traps

TRAP TYPES CAUSESStructural Traps

Fold Traps: Compressional Folds Compactional Folds Diapir Folds

Tectonic processes Depositional / Tectonic processes Tectonic Processes

Fault Traps Tectonic Processes

Stratigraphic Traps Depositional morphology or diagenesis

Hydrodynamic Traps Water flow

Combination Traps Combination of two or more of the above processes

Structural traps are primarily due to post-depositional processes which modify the spatial configuration of the reservoir rock, mainly by folding and faulting.

Stratigraphic traps are those whose geometry is due to changes in lithology. The lithological changes may be depositional, as in channels, reefs and bars, or post-depositional, where strata are truncated or where rock lithologies have been altered by diagenesis.

In hydrodynamic traps, the downward movement of formation waters prevents the upward movement of oil. Combination traps combine two or more of the previously-defined generic groups. A good summary of the more common trap types and their formational environments is found in Bailey and Stoneley (1981).

Structural Traps

The geometry of a structural trap is due essentially to some post-depositional modification of the reservoir.

In the words of Levorsen (1967) "A structural trap is one whose upper boundary has been made concave, as viewed from below, by some local deformation, such as folding, or faulting, or both, of the reservoir rock."

Selley (1982) has further defined the boundaries of a structural trap, thus "The edges of a pool occurring in a structural trap are determined wholly or in part by the intersection of the underlying water table with the roof rock overlying the deformed reservoir rock." Structural traps are divided into those due to folding, and those due to faulting.

Fold Traps ( Compressional )

Anticlinal traps which are due to compression are most likely to be found in or near geosynclinal troughs. These troughs are usually associated with active continental margins where there is a net shortening of the earth's crust ( Figure 1 , Active continental margin with net shortening of crust- subduction zone).

Figure 1

In California, the Tertiary basins form a major hydrocarbon province which contains compressional anticlinal traps. Within this province are a number of fault-bounded troughs infilled by thick regressive sequences in which organic-rich basinal muds are overlain by deep-sea sands and capped by younger continental beds as shown by Figure 2 (Generalized west-southwest-east-northeast structural cross-section), a cross- section of the Los Angeles basin.

Figure 2

These basins have been locally subjected to tight compressive folding associated with the apparent transcurrent movement of the San Andreas fault system (Barbat, 1958; Schwade at al., 1958; and Simonson, 1958). Anticlinal traps of a broad, gentle character may also be formed in large cratonic basins of stable shelf sediments. Many oil and gas fields in this province are also associated with faulting, either normal, reverse or strike-slip.

The Wilmington oil field in the Los Angeles basin ( Figure 3 , Oil fields of the Los Angeles basin) is a giant anticlinal trap with ultimate recoverable reserves of about 3 billion barrels of oil.

Figure 3

It is approximately 15 kilometers long and nearly 5 kilometers wide. The overall anticlinal shape of the field is shown by the structure contours on top of the main pay zone ( Figure 4 , Structural contours on top of Ranger zone, Wilmington field, CA). Notice also the cross-cutting faults.

Figure 4

From a southwest-northeast cross section of the Wilmington field, you can see the broad arch of the anticline ( Figure 5 , Southwest-northeast cross-section A-Z, Wilmington field). The main reservoir occurs beneath the Pliocene unconformity in Miocene- and Pliocene-age deep-sea sands.

Figure 5

The foothills of the Zagros mountains in Iran contain one of the best-known hydrocarbon provinces with production from compressional anticlines ( Figure 6 , Location map, southwest Iran and Persian Gulf).

Figure 6

Individual anticlines are up to 60 kilometers in length and 10-15 kilometers in width. Sixteen of these anticlinal fields are in the "giant" category with recoverable reserves of over 500 million barrels of oil or 3.5 trillion cubic feet of gas (Halbouty et al., 1970). The Asmari limestone (Oligocene-Miocene) , a reservoir with extensive fracture porosity, provides the main producing reservoir. Some single wells have flowed up to 50 million barrels. Figure 7 (Southwest-northeast generalized sections through Asmari oil fields)

Figure 7

shows two schematic cross sections through the Asmari oil fields according to two different interpretations of deep structure; one showing anticlines without thrusting and one with thrust faulting. For further detailed descriptions of these fields, the reader is also referred to Lees (1952), Falcon (1958, 1969) and Colman-Sadd (1978).

In areas of still more intense structural deformation, anticlinal development may be associated with thrust faulting. Such thrust fault belts are usually found within mountain chains throughout the world. The thrust faults cause a thickening of the sedimentary column as older rocks are thrust up over younger rocks causing repeated sections. Traps may occur in anticlines above thrust planes, and in reservoirs sealed beneath the thrust.

In Wyoming, the Painter Reservoir field is a fairly tight anticline ( Figure 8 , Structural contours on top of Nugget sandstone, Painter Reservoir field, Wyoming) beneath a thrust plane, which itself is involved in thrusting along its southeastern border.

Figure 8

In cross section, the anticline is overturned and thrust faulted on its southeastern flank ( Figure 9 , Northwest-southeast cross-section through Painter Reservoir field). The anticline occurs beneath a series of thrust slices that in turn occur beneath a major unconformity.

Figure 9

Fold Traps ( Compactional )

Compactional fold frequently occurs where crustal tension associated with rifting causes a sedimentary basin to form. The floor is commonly split into a system of basement horsts and grabens. An initial phase of deposition fills this irregular topography. Anticlines may then occur in the sedimentary cover draped over the structurally-high horst blocks ( Figure 1 , Compactional anticlines in sediments draped over underlying structurally high horst blocks ).

Figure 1

These anticlines develop by differential compaction of sediment. At the time of deposition, thickness of a given sedimentary unit is thinner over the crest of the underlying structural high ( Figure 2a , Developmental stages of compactional anticlines--initial stage of deposition).

Figure 2a

Compaction then takes place over the feature ( Figure 2b , Developmental stages of compactional anticlines--compactional stage). Though the percentage of compaction is constant for crest and trough, the actual amount of compaction is greater for the thicker sediment in the trough. Deep-seated, recurrent fault movement may enhance the structural closure ( Figure 2c , Developmental stages of compactional anticlines--structural closure enhanced by recurrent fault movement).

Differential depositional rates may also enhance the closure. Carbonate sedimentation tends to be thicker in the shallower waters over underlying structural highs. Therefore, shoal and reefal facies may pass off-structure into thinner increments of basinal lime mud. Sandbar or shoal sands may also develop on the crest of structures, with deep-water muds present further down the flanks. For this reason, reservoir quality often diminishes down the flank of such structures.

In the North Sea there are several good examples of compactional anticline traps where Paleocene deep-sea sands are draped over deep-seated basement horsts. These include the Forties (Hill and Wood, 1980), Montrose (Fowler, 1975), and East Frigg fields (Heritier et al., 1980).

The Forties field is an example of a compactional anticline on the western side of the North Sea. Regional structure is a southeasterly-plunging nose bounded to the

northeast and southwest by faults ( Figure 3 , Structural contours on top of Paleocene reservoir, Forties field area, North Sea).

Figure 3

A north-south cross section depicts the anticline developed at the Paleocene level where the reservoir sands are sealed by overlying Tertiary clays ( Figure 4 , Schematic north-south cross-section A-Z through Forties field, North Sea).

Figure 4

The anticline overlies a deep-seated horst of late Jurassic volcanics. Source rocks of upper Jurassic age occur around the edge of this horst structure. Differential compaction and recurrent fault movement seem to have controlled the structure throughout the Cretaceous and into the Tertiary.

Only differential compaction folds occurring over deep-seated horst blocks have been discussed. Compaction folds, however, may also occur over reefs and other deep-seated rigid features.

Fold Traps; Comparison of Major Types

There are many differences between the fold traps caused by compression, and those caused by compaction (Selley, 1982). Compressional folds form after sedimentation, so the porosity found in them is more related to primary, depositional causes than to structure. These folds may also contain fracture porosity as they are usually lithified when deformed.

With compaction folds, porosity may vary between crest and flank. As already discussed, there may be primary depositional control of reservoir quality. Furthermore, secondary diagenetic porosity may also be developed on the crests of compactional folds as such structures are prone to sub-areal exposure and leaching.

Compressional folds are generally oriented with their long axis perpendicular to the axis of crestal shortening, whereas compactional folds are often irregularly shaped due to the shape of underlying features.

Compressional folds commonly form from one major tectonic event, while compactional folds may have a complex history due to rejuvenation of underlying basement faults.

Depositional Traps

Stratigraphic trap geometry is due to variations in lithology. These variations may be controlled by the original deposition of the strata, as in the case of a bar, a channel or a reef. Alternatively, the change may be post-depositional as in the case of a truncation trap, or it may be due to diagenetic changes.

For reviews on the concept of stratigraphic traps, the reader is referred to Dott and Reynolds (1969) and Rittenhouse (1972). Major sources of specific data on stratigraphic traps can be found in King (1972), Busch (1974), and Conybeare (1976).

Levorsen (1967) defines a stratigraphic trap as "one in which the chief trap-making element is some variation in the stratigraphy, or lithology, or both, of the reservoir rock, such as a facies change, variable local porosity and permeability, or an upstructure termination of the reservoir rock, irrespective of the cause."

Stratigraphic traps are harder to locate than structural ones because they are not as easily revealed by reflection seismic surveys. Also, the processes which give rise to them are usually more complex than those which cause structural traps.

A broad classification of the various types of stratigraphic traps can be made. However, classifying traps has its limitations because many oil and gas fields are transitional between clearly-defined types.

Table 1 ,

Table 1

(Classification of stratigraphic type hydrocarbon traps) based on the scheme proposed by Rittenhouse (1972), shows that a major distinction can be made between stratigraphic traps which occur within normal conformable sequences ( Figure 1 ,

Figure 1

Schematic of trap within normal conformable sequence) and those that are associated with unconformities ( Figure 2 , Schematic of traps associated with unconformaties).

Figure 2

This distinction is rather arbitrary since there are some types, such as channels, that can occur both at unconformities and away from them ( Figure 3 , Schematic of two channel traps).

Figure 3

Of the traps occurring within normal conformable sequences, a major distinction is made between traps due to deposition and those due to diagenesis. The depositional or facies-change traps include channels, bars and reefs.

Depositional Traps: Channels

Many oil and gas fields occur trapped within channels of various types, ranging from meandering fluvial deposits through deltaic distributary channels to deep-sea channels.

Many good examples of stratigraphic traps in channels can be found in the Cretaceous basins along the eastern flanks of the Rocky Mountains, from Alberta, through Montana, Wyoming, Colorado and New Mexico. These channels occur both cut into a major pre-Cretaceous unconformity and within the Cretaceous strata.

The South Glenrock oil field in Wyoming contains oil trapped in both marine-bar and fluvial-channel reservoirs. The channel reservoir has a width of some 1500 meters and a maximum thickness of approximately 15 meters ( Figure 1 , Isopach map of Lower Muddy interval, South Glenrock oil field, Wyoming). It has been mapped for a distance of over 15 kilometers and can be clearly seen to meander.

Figure 1

A cross section of the field shows that the channel is only partially filled by sand and is partly plugged by clay ( Figure 2 , West-east cross-section A-Z of two Lower Muddy stream channels).

Figure 2

The SP curves on some of the well logs (e.g. wells #5 and #6 on Figure 2 ) display upward-fining point-bar sequences, a characteristic of meandering channel deposits.

The South Glenrock field illustrates an important points about channel stratigraphic traps. Because of their limited areal extent and thickness, such reservoirs seldom contain giant accumulations.

The deltaic distributary channel of Oklahoma, is a good example of channel traps in sands other than the meandering fluvial variety ( Figure 3 , Isopach map of Booch sandstone, greater Seminole district, eastern Oklahoma).

Figure 3

Depositional Traps: Bars

Because of their clean well-sorted texture, marine barrier bars often make excellent reservoirs (Hollenshead and Pritchard, 1961).

The barrier sands may coalesce to form blanket reservoirs. Oil may then be structurally or stratigraphically trapped within these blanket sands. Sometimes, however, isolated barrier bars may be totally enclosed in marine or lagoonal shales, forming stratigraphic traps in shoestring sands elongated parallel to the paleo shoreline ( Figure 1 , Schematic of barrier bars, showing interconnedted bars forming blanket reservoir and one isolated bar set).

Figure 1

The Rocky Mountain Cretaceous basins contain many barrier bar stratigraphic traps. The Bisti field in the San Juan basin, New Mexico is a classic barrier bar sand (Sabins, 1963, 1972). The field is about 65 kilometers long and 7 kilometers wide ( Figure 2 , Bar sandstone isopach map of Bisti field, Colorado).

Figure 2

It consists of three stacked sandbars, with an aggregate thickness of 15 meters, totally enclosed in the marine Mancos shale ( Figure 3 , North-south cross-section A-Z of Bisti field using electric logs).

Figure 3

The SP log in some wells shows the typical upward-coarsening grain-size motif which characterizes barrier bars. (See inset, Figure 3 .) Two other examples of barrier bar stratigraphic traps are the Bell Creek field, Montana (Berg and Davies, 1968; McGregor and Biggs, 1970, 1972); and the Recluse field, Wyoming (Woncik, 1972).

During a regressive stage, barrier bars often develop as sheet sands, which may pass updip into lagoonal or intertidal shales causing pinch-out or feather-edge traps (Selley 1982). As with many sheet reservoirs, lateral closure must occur for the trap to be valid. This may be stratigraphic, as for example, where an embayment occurs in a shoreline. Alternatively, it may be structural, in which case the trap might be more properly classified as a combination trap (Selley, 1982).

Depositional Traps: Reefs

The reef or carbonate build-up trap has a rigid stoney framework containing high primary porosity (Maxwell, 1968; Jones and Endean, 1973). Reefs grow as discrete domal or elongated barrier features, and have long been recognized as one of the most important types of stratigraphic traps.

Reefs are often later transgressed by organic-rich marine shales (which may act as source rocks) or the reefs may be covered by evaporites. Oil or gas may be trapped

stratigraphically within the reef, with the shales or evaporites providing excellent seals.

In Alberta, Canada, the Devonian-age Rainbow reefs in the Black Creek Basin provide an excellent example of reef traps (Barss et al., 1970). More than seventy individual reefs, containing various amounts of oil and gas, were discovered within an area about 50 kilometers long and 35 kilometers wide. Total reserves of these reefs are estimated in excess of 1.5 billion barrels of oil in place and one trillion cubic feet of gas.

As shown in Figure 1 (Schematic cross-section through Middle Devonian reefs, Rainbow area, Alberta, Canada), two basic geometric forms of reefing are present: the pinnacle reef and the broader elliptical form of the atoll reef.

Figure 1

The individual reefs are up to 15 square kilometers in area and up to 250 meters high in relief. At the end of reefal growth, evaporite sediments infilled the basin. The evaporites completely covered the reefs, thereby providing an excellent seal for hydrocarbon entrapment.

There is a wide range of net pays found in the Rainbow reefs ( Figure 1 ). Some reefs are nearly full of oil and gas, while others contain a very small column of oil or gas at the very crest of the reef. Porosities and permeabilities also differ greatly from reef to reef as well as within individual reefs. Such changes are due to variations in lithofacies and diagenetic effects, and are typical features of reefal traps ( Figure 2 , Cross-section of pinnacle reef showing complex lithofacies,Rainbow area, Alberta, Canada).

Figure 2

There are many other reef hydrocarbon provinces around the world, notably in the Arabian Gulf and Libya. In Libya, the Intisar reefs in the Sirte basin have been well documented (Terry and William, 1969; Brady et al., 1980).

Diagnenetic Traps

Diagenetic traps are formed by the creation of secondary porosity in a non-reservoir rock by replacement, solution or fracturing with the tight unaltered rock forming the seal for the trap (Rittenhouse, 1972).

An example of a diagenetic trap formed by replacement is the Deep River field in Michigan, in which dolomitization of a preexisting limestone deposit has resulted in the formation of secondary intercrystalline porosity ( Figure 1 ).

Figure 1

The development of solution porosity is commonly associated with carbonate rocks ( Figure 2 ), but may involve sandstones as well.

Figure 2

Fracturing can cause secondary porosity in any brittle rock — whether carbonate, sandstone, shale, igneous or metamorphic rock (Kostura and Ravenscroft, 1977). The Spraberry trend in west Texas forms a series of diagenetic traps (with oil reserves of about one billion barrels) within a producing fairway about 240 kilometers long and 80 kilometers wide (Wilkinson 1953). A structure map contoured on the productive Spraberry formation, a 300-meter-thick section of tight Middle Permian shales, siltstones, limestones, and fine-grained sandstones shows that in the southern Midland basin, the areas of oil production have little relationship to structure ( Figure 3 ). Production comes from areas of fracturing throughout the otherwise tight Spraberry formation.

Figure 3

The depositional and diagenetic stratigraphic traps just considered occur in normal comformable sequences, although they may also occur at unconformities.

UNCONFORMITY-RELATED TRAPSAnother major group of stratigraphic traps is represented by traps for which an unconformity is essential (Table 1 ) (Levorsen, 1934).

Table 1

Significantly large percentages of the known global petroleum reserves are trapped adjacent to major unconformities. In addition to being held in pure stratigraphic traps, many of these reserves are held in structural and combination traps as well. Unconformity-related traps can be subdivided into those which occur above the unconformity and those beneath (Figure 1 , Schematic of traps located above and below an unconformity).

Figure 1

Traps which occur above an unconformity will be discussed first.

Shallow-marine or fluvial sands may onlap a planar unconformity. A stratigraphic trap can occur where such sands are overlain by shales and are underlain by impermeable rock which provides a seat seal. Onlapping updip pinch-out sands such as these could occur as sheets (Figure 2a , Schematic of onlapping pinch-out sands-- occurring as a sheet deposit) , or as discrete paleogeomorphic traps (Figure 2b , Schematic of onlapping pinch-out sands--occurring as a discrete paleogeomorphic sand).

Figure 2a, 2b

A good example of an onlap stratigraphic trap is provided by the Cut Bank field of Montana, with recoverable reserves of over 200 million barrels of oil (MacKenzie, 1972). Here the Cretaceous Cut Bank sand unconformably onlaps Jurassic shales and is itself onlapped by younger shales (Blixt, 1941; Shelton, 1967). Figure 3 , (Southwest-northeast E-log correlation section A-Z, Cut Bank sandstone, Montana) is a cross section through this field.

Figure 3

One type of paleogeomorphic trap is represented by channels which cut into the unconformity; another occurs where sands are restricted within strike valleys cut into alternating hard and soft strata (Figure 4 , Schematic of channel and strike valley sands above an unconformity) (Harms, 1966; Martin, 1966; and McCubbin, 1969).

Figure 4

It is important to note that closure is necessary along the strike of such traps, not just updip as shown in Figure 2a . In Figure 5 (Schematic of sandstone pinch-out intersecting with a structural nose), closure is provided by the intersection of a sandstone pinch-out with a structural nose.

Figure 5

The second group of traps associated with unconformities is truncation traps which occur beneath the unconformities (Figure 6 , Schematic of traps below unconformity).

Figure 6

Again, it is generally overlying shales which provide a seal (and often the source as well) for such traps. As with onlap, pinch-out, and paleogeomorphic traps, closure is needed in both directions along the strike (Figure 7 , Schematic of trap below unconformity, featuring closure provided by the intersection of a dipping structural nose and a flat unconformity).

Figure 7

This may be structural or stratigraphic but for many truncation traps, it may be provided by the irregular topography of the unconformity itself, such as a buried hill providing closure for a subcropping sandstone formation (Figure 8 , Schematic of trap below unconformity, featuring closure provided by buried hill).

Figure 8

Many truncation traps have had their reservoir quality enhanced by secondary solution porosity due to weathering. Secondary solution porosity induced by weathering is most common in limestones, but also occurs in sandstones and even basement rock. Examples in limestones are found in Kansas, and in the Auk field of the North Sea (Brennand and van Veen, 1975). Development of subunconformity solution porosity in sandstones has occurred in the Brent Sand of the North Sea (Bowen, 1972), and in the Sarir group of Libya (Selley 1982). Basement rock weathering is found in the Augila field of Libya (Williams 1968, 1972). One of the best known truncation traps in the world is the East Texas field (Halbouty, 1972; Halbouty and Halbouty, 1982) which contained over 5 billion barrels of recoverable oil. The trap is caused by the truncation of the Cretaceous Woodbine sand by the overlying impermeable Austin chalk (Figure 9 , Generalized west-east cross-section, East Texas basin).

Figure 9

It has a length of some 60-70 kilometers and a width of nearly ten kilometers. Figure 10 (Structural contours on top of Woodbine sand, East Texas field) illustrates the structural closure at the northern and southern ends of the field.

Figure 10

Hydrodynamic Traps

In a hydrodynamic trap, a downward movement of water prevents the upward movement of oil or gas. Pure hydrodynamic traps are extremely rare, but a number of traps result from the combination of hydrodynamic forces and structure or stratigraphy.

An ideal hydrodynamic trap is shown in Figure 1 (Schematic cross-section of an ideal hydrodynamic trap).

Figure 1

A monoclinal flexure is developed which has no genuine vertical closure; oil could not be trapped within it in a normal situation. Groundwater, however, is moving down through a permeable bed and is preventing the upward escape of oil. Oil is trapped in the monoclinal flexure above a tilted oil-water contact. Pure hydrodynamic traps like this, however, are very rare.

There are a number of fields with tilted oil-water contacts where entrapment is a combination of both structure and hydrodynamic forces ( Figure 2 , Schematic cross-section showing entrapment from both structural and hydrodynamic forces).

Figure 2

For further discussion of the effect of hydrodynamic conditions on hydrocarbon traps, the reader is referred to Goebel (1950), Hubbert (1953), Yuster (1953), and Eremako and Michailov (1974).

Combination Traps

Combination traps result from two or more of the basic trapping mechanisms ( structural, stratigraphic, and hydrodynamic ). Since there are many ways in which combination traps can occur, a few examples must suffice for explanation.

In the Main Pass Block 35 field of offshore Louisiana, a rollover anticline has developed to the south of a major growth fault (Hartman, 1972) ( Figure 1 , Structural contours on top of 'G2' sandstone, Main Pass Block 35, offshore Louisiana).

Figure 1

The rollover anticline, however, is crosscut by a channel. Oil with a gas cap occurs only within the channel; thus, the trap is due to a combination of structure and stratigraphy.

An excellent example of a combination trap is provided by the Prudhoe Bay field on the North Slope of Alaska (Morgridge and Smith, 1972; Jones and Speers, 1976; Jamison et al., 1980; Bushnell, 1981). A series of Carboniferous-through-basal-Cretaceous strata were folded into a westerly-plunging anticlinal nose ( Figure 2 , Structural contours on top of Sadlerochit reservoir, Prudhoe Bay, Alaska).

Figure 2

This nose was truncated progressively from the northeast, and overlain by Cretaceous shales which acted as source and seal to the trap. Oil and gas were trapped in reservoir beds subcropping the unconformity, primarily in the Triassic Sadlerochit sandstone. Major faulting on the northern and southwestern side of the structure provided additional closure.

Fault-unconformity combination traps characterize the northern North Sea. Jurassic sandstone reservoirs exist in numerous tilted fault blocks which were truncated and overlain by Cretaceous shales. The resulting traps include such fields as Brent (Bowen, 1972), Ninian (Albright et al., 1980), and Piper (Maher, 1980). A cross section through one of these, the Piper field, is shown in Figure 3 (Southwest-northeast structural cross-section, Piper field, North Sea).

Figure 3

Diapir Associated Traps

Diapirs are a major mechanism for generating many types of traps. Diapirs are produced by the upward movement of less dense sediments, usually salt or overpressured clay. Recently-deposited clay and sand have densities less than salt which has a density of about 2.16 g/cm3.

As most sediments are buried, they compact, gaining density; ultimately, a depth is reached where sediments are denser than salt. This generally occurs between 800 and 1200 meters. When this situation is reached, the salt tends to flow upwards to displace the denser overburden. If this movement is triggered tectonically, the resulting structure may show some alignment, such as that displayed by the salt domes in the North Sea ( Figure 1 , Salt structures of the southern North Sea). However, in many cases, the salt movement is apparently initiated at random.

Figure 1

Movement of salt develops several structural shapes, from deep-seated salt pillows which generate anticlines in the overlying sediment, to piercement salt domes which actually pierce the overlying strata ( Figure 2 , Schematic cross-section showing two salt structures; a salt pillow on the right and a piercement salt dome on the left) (Bishop, 1978). In extreme cases, salt diapirs can actually penetrate to the surface as in Iran (Kent, 1979).

Figure 2

There are many ways in which oil can be trapped on or adjacent to salt domes (Halbouty, 1972) ( Figure 3 , Schematic cross-section showing the varieties of hydrocarbon traps associated with piercement salt domes).

Figure 3

There may be simple structural anticlinal or domal traps over the crest of the salt dome. Notable examples of this type include the Ekofisk field (Van der Bark and Thomas, 1980), and associated fields of offshore Norway and Denmark. There may also be complexly-faulted domal traps, stratigraphic pinch-out and truncation traps , or unconformity truncation traps.

Occasionally anticlinal structures known as turtle-back structures are developed between adjacent salt domes. When the salt moves into a dome, the source salt is removed from its flanks, thereby developing rim synclines. Thus, anticlines develop above the remaining salt ( Figure 4 , Schematic cross-section showing a turtleback structure (anticline) developed between two adjacent piercement salt domes). The Bryan field of Mississippi is an example of a turtle-back trap (Oxley and Herling, 1972).

Figure 4

Major oil and gas production from salt-dome-related traps comes from the U.S. Gulf Coast, Iran, the Arabian Gulf and the North Sea.

Diapiric mud structures, not just salt domes, may also generate hydrocarbon traps. Sometimes diapirs of overpressured clay intrude the younger, denser cover and, just like salt domes, mud lumps may even reach the surface.

Frontier vs. Mature Basins

There are two general exploration situations that influence prospect generation. The first situation involves frontier, or less mature basins, where the presence of hydrocarbons has not been established. Here, our main concern is whether hydrocarbons have matured and been expelled from a source rock. The second situation involves more mature basins where we know a hydrocarbon source exists. Here, our main concerns include determining migration pathways, the types and occurrences of reservoirs, and the effectiveness of seals. For either situation, we must consider the impact of the types, quantity, and quality of data available for answering these questions.

Expert systems are designed to mimic, quantify and analyze the objective and subjective decision-making processes of humans. Analysis of decision-making processes allows us to quantitatively and qualitatively determine how explorationists generate prospects. Analysis of how explorationists evaluate prospects enables us to predict the potential impact of basin models on decisions about whether to drill a given exploration concept. Newly developed quantitative stratigraphic models show how quantitative models strongly affect the outcome of the decision-making process.

Estimating geologic processes, configurations and lithologic distributions is essential in prospect generation, and predicting lithology is perhaps the weakest link. Predicting lithology is especially difficult in many structural plays and in stratigraphic plays that lack an associated seismic anomaly. Stratigraphic and basin-fill models will significantly affect prospect generation because they offer more accurate technologies for predicting subsurface lithology.

Impact of Stratigraphic Models

Before generating a prospect, explorationists commonly believe that the prospect exists. This prior belief may be based on drilling history, statistical analysis or analog reasoning. Explorationists modify their prior belief based on the amount and quality of available data relating to the prospect. They may further modify their belief as new data become available. For example, a new seismic line may give favorable, but not definitive, information about the existence of an anticline ( Figure 1 , Modification of a prior belief after new information is introduced.

Figure 1

In Bayes Theorem, the final odds that a prospect exists given seismic information are equal to the likelihood of the prospect given seismic data multiplied by the prior odds that the prospect exists.). We express probabilities as subjective degrees of belief. At

any given time, there is a probability that an event will occur. As new evidence is introduced, we perceive the original probability to increase, decrease, or remain the same.

The extent to which we modify our belief in the prior existence of a particular prospect depends on the Predictiveness and favorableness of evidence for the prospect. Predictiveness relates to the sense and magnitude of change in belief that occurs when new evidence is introduced. Favorableness depends on Predictiveness and relates to the degree of certainty that the evidence exists or is optimally developed for a specific example. Predictiveness is an abstract property that relates to the certain existence or optimal development of evidence in general, while favorableness depends on its quality or uncertainty of existence for a particular situation.

The rule-based expert system GEORISKTM, developed at the Colorado School of Mines, uses Bayes Theorem to quantify the Predictiveness and favorableness of geological, geophysical and geochemical information for prospect evaluation (Lessenger, 1988a, 1988b). The system was built by quantifying the decision-making processes of three exploration managers. Rules are in the form of hypothesis updates that have premise, action and certainty factors:

Premise IF Evidence

Action THEN Hypotheses (Certainty Factors)

Within GEORISKTM, hypotheses are predictions of causal processes and configurations, or assessments of information that we can use to infer causal processes and configurations. Evidence can be observed data or inferences from data. Certainty factors quantify the predictiveness of positive and negative evidence for the hypotheses, and can range from -100 to +100. For example, consider the following GEORISKTM rule:

Premise IF Thermal history is (is not) well-constrained

Action THEN Thermal model is accurate +90 (-90)

A prior belief that the thermal model is accurate is modified with a certainty factor of +90 if the thermal history is perfectly constrained and -90 if it is not at all established. We assign ranges in values between +90 and -90 according to the degree to which we consider the thermal history constrained.

According to Bayesian analysis, the greater the predictiveness of evidence, the more the prior belief will be modified and will result in a new belief or outcome (Tversky and Kahneman, 1982). The magnitudes of certainty factors control the amount a prior belief is modified. We should therefore look at the spread in certainty factors for favorable and unfavorable evidence to qualify the perceived predictiveness of evidence.

The impact of favorable evidence is termed the sufficiency and the impact of unfavorable evidence is termed the necessity for an hypothesis ( Figure 2 , The necessity and sufficiency of evidence for a hypothesis.

Figure 2

The spread in the necessity and sufficiency of evidence controls the spread in potential certainty values for the hypothesis. This is a measure of the Predictiveness of evidence for a hypothesis.). To prove an hypothesis, both the sufficiency and necessity must be high enough to increase confidence that the hypothesis is true. For example, the predictiveness of thermal history constraints for an accurate thermal model is qualitatively high (D=180); thermal history constraints are both highly necessary and highly sufficient to ensure thermal model accuracy. If the sufficiency and necessity have magnitudes approaching zero, our belief in the hypothesis is modified very little with either positive or negative evidence, and we can not prove the hypothesis.

Many sources of information are available for prospect generation, including seismic data, well logs, maturation and geodynamic models, core analyses, facies models, analog field studies, etc. There are three types of information: 1) observational, 2) inferential and 3) quantitative. Observational information requires only objective measurement; interpretation and inference is minimal (e.g., TOC values). Inferential information requires a conceptual interpretation of data (e.g., depositional environment). Quantitative information requires manipulation of data based on descriptions of the data (e.g., an isopach map) or on a quantitative model (e.g., a McKenzie stretching model) rather than a conceptual model (e.g., a facies model). Seismic data supply explorationists with all three types of information: delineating seismic faults is observational; interpreting seismic facies based on reflectors within a depositional sequence is inferential; and converting seismic times to depth based on a velocity model is quantitative.

Different types of information are perceived as having different degrees of Predictiveness for either the same source of information or for predicting the same causal process. Information perceived as highly predictive has a greater impact on modifying the prior belief than does information perceived as less predictive. Quantitative methodologies are perceived as most predictive, inferential methodologies least predictive, and observational methodologies of intermediate Predictiveness. Predictiveness also increases when multiple types of information mutually predict the same causal process or favorable condition.

For example, to predict a favorable structural configuration, quantitative depth conversion (D=120) is more predictive than observational seismic data quality (D=110) ( Figure 3 , Perceived relative Predictiveness of different types of information used to assess structural configuration.).

Figure 3

Both are more predictive than a structural cross section (D=80), an interpretational inference from well logs, even though well logs imply drilling and therefore larger amounts of, and potentially more reliable, information near the prospect. We often perceive well logs as providing more direct, and thus more reliable, information for predicting the presence and magnitude of faulting than seismic data. But because well logs lack lateral information and lateral changes in structural configuration can be dramatic, we actually perceive seismic data as more reliable for predicting faulting than well logs. The reliability of information depends on how we use it. Consequently, indirect information may be considered more reliable than direct, despite the opposite conclusion reached through intuition.

Maturation and source potential analyses have large spreads in necessity and sufficiency (D=180), and may be perceived as very predictive of the presence of hydrocarbons on a basin-wide scale ( Figure 4 , Perceived relative power and Predictiveness of different types of information, representing three major elements used to assess a prospect.).

Figure 4

These analyses are highly quantitative, involving, for example, thermal maturation modeling and geohistory estimates. The evidence and methodologies assessed increase, both in number and variety, as compared to other analyses (e.g., estimating the distribution of reservoir lithology).

Structural analyses depend primarily on the observation of data, particularly seismic. There is a moderate spread in the necessity and sufficiency (D=100) of structural information for predicting the presence of a hydrocarbon trap. The number and variety of available analyses is less than is available for maturation studies. While structural and stratigraphic analyses are equally necessary for predicting the existence of a prospect (-50), structural analyses are far more sufficient than stratigraphic analyses (+50 compared with +20).

Methods for predicting lithology, in contrast, rely on inferential data types, unless there is a seismic anomaly associated with reservoir development. The presence of lithologic conditions for predicting a hydrocarbon reservoir and trap has a relatively low necessity and sufficiency spread (D=70).

The methodologies we use to predict lithology depend on specialized conditions. If we have a well nearby (i.e. within one mile) that indicates the presence of the target lithology, predictiveness increases. If there is a seismic anomaly, predictiveness increases further, and if additional sources of information, such as well ties to seismic

data and models, are available, predictiveness is maximized. As with structural analyses, there are different degrees of direct information for predicting lithology: core data are most direct, well logs less direct and seismic anomalies least direct. But, seismic data contain lateral information and are therefore more predictive of lithology than either core or well logs, even though both core and well logs are more direct measurements of lithology. As with structural analyses, direct measurements are not necessarily the most predictive.

We often do not have a nearby well or seismic anomaly. Our lithologic analyses are reduced to evaluations of regional or sub regional trends that, when combined with the low certainty levels in trend data, can substantially reduce the effective necessity and sufficiency of predicting lithologic conditions. Without seismic anomalies and nearby wells, we have only very weak methods for predicting the lateral distribution of specific lithology from well and core information. Currently, these weak methods are derived from concepts or empirical constructions, such as facies models, that are non-quantitative, weakly predictive, and that provide no estimate of the accuracy or confidence level of their prediction.

At present, there is a large imbalance in the amounts and types of information available for predicting the necessary conditions for a hydrocarbon accumulation. Seismic information is widely available and is highly predictive of structural configurations. Consequently, predicting structural traps has a greater certainty than predicting stratigraphic traps or reservoir facies distribution. A variety of information sources exist for predicting source, maturation and migration. Although source-maturation-migration technologies are not completely predictive of hydrocarbon availability, their use has greatly increased predictiveness.

The amounts, types and perceived qualities of information available for predicting the distribution of specific lithology are limited, and our ability to predict lithology is low. In the common situation where we have neither a stratigraphic seismic anomaly nor nearby wells, predicting the distribution of lithology is the weakest link in prospect generation. The situation is somewhat analogous to the prediction of hydrocarbon availability before source-maturation-migration technologies were introduced and we relied primarily on play-specific, empirical, non-quantitative models. Without seismic anomalies, predicting lithology currently relies on empirical trends and conceptual facies models.

Impact of Stratigraphic Models

Stratigraphic and basin-fill models will significantly affect prospect generation decisions because they simulate what is perceived as reliable information and because of their potential for increasing our ability to predict lithology.

Three conditions affect the reliability of information for prospect generation. First, multiple information types that support the same predictions are more reliable than a single information type, even if that single type is a strong one. Due to the sparsity of information types currently used in lithology prediction, the future introduction of new quantitative methods, such as those developed through the use of 3-D seismic data, will have a significant impact. In addition to adding a quantitative information type, other analyses necessary for modeling -- such as correlation theory, sequence analysis, volumetric facies partitioning, stratigraphic simulation inversion, geohistory and geodynamics -- will also add to the types and sources of information available for constraining predictions of lithology.

Second, reliable information must be consistent with other types of information. Stratigraphic models require coherent information from many different types and sources of information because they incorporate many different kinds and scales of controlling processes. For example, a simulation model must be consistent with seismic data and well logs. Separating eustatic from subsidence variations requires knowledge of synchronous stratigraphic sequences in different geographic locations. We can constrain subsidence variations with modeling of regional geodynamic processes and plate tectonic configurations. Sediment supply and dispersal may require knowledge of gross climatic conditions, basin configurations and sedimentary processes. Because we run basin-fill models in time and not depth, methods for extracting and constraining temporal relationships such as biostratigraphy, age-dating and physical sequence stratigraphy are critical. These separate evaluations must all be coherent and geologically reasonable. Coherent, multiple information types increase our confidence in the model results and positively modify our prior belief.

Third, we perceive quantitative models as more predictive than conceptual models because they allow us to quantitatively relate processes to responses and to explicitly state assumptions and initial and boundary conditions, allowing us to establish degrees of confidence in predictions. Quantitative models are necessary for understanding the sensitivity of the stratigraphic response to changes in processes. Geologic assumptions and biases are explicit in quantitative models. By building, testing and using quantitative models, we commonly uncover unreasonable assumptions that were previously not obvious. Sensitivity analyses and explicit statements of assumptions and biases allow us to estimate degrees of non-uniqueness and confidence in model predictions. In contrast, non-quantitative models require only conceptual inferences of responses. Assumptions and boundary and initial conditions are not explicitly stated. Therefore, we cannot reliably estimate degrees of non-uniqueness, nor our confidence in model predictions.

We use only two sources and two types of information in predicting lithology. One source, well logs, is inferential, so we perceive this source as having low predictiveness for estimating the distributions of specific lithology. The other source, seismic data, is primarily inferential, except when we use quantitative seismic models to increase our confidence in the geologic significance of waveform anomalies. If within a particular play, seismic anomalies prove to be reliable, we perceive them as highly predictive. Unfortunately, seismic anomalies for predicting lithology are rare. By incorporating stratigraphic models, the perceived value and ability to predict lithology (i.e., alter a prior belief) will be increased, at least to a level similar to predictions of structure and hydrocarbon presence.

Because we may perceive stratigraphic models as highly predictive, we may also expect highly accurate predictions. Stratigraphic models are in their infancy and are only now being built and tested. We need to calibrate models to stratigraphic data to determine what conditions models are applicable to, and to estimate the confidence limits in predictions, given the approximations and measurement errors inherent in the model and the sparse sampling of stratigraphic data. At this stage of model development, we need an iterative process of model building, testing and rebuilding to gain confidence in model predictions.

Anomaly

Prospect generation begins with the search for anomalies. An anomaly may be defined as a deviation from whatever trend is normal. It's a local feature that can be

distinguished within a larger area, because it has some kind of distinctive fingerprint which makes it stand out from the background data. The anomaly can be revealed by geologic mapping, geophysical or geochemical data, by biological and soil surveys, or by anything that departs from the norm.

The petroleum geologist knows that some anomalies are associated with deposits of commercially valuable oil and gas. The usual trend is for subsurface petroleum to make its way to the surface and eventually dissipate. In order to prevent further migration, there must be an anomaly present to act as a barrier. Geologists often regard anomalies as being broadly synonymous with structure, but anomalies are also associated with stratigraphic and other trap types.

Often, more than one theory may account for an anomalous geological situation. If our aim is to generate the maximum number of prospects, it makes sense to use the most fruitful theory.

Lead The likelihood of a commercial accumulation of hydrocarbons can be increased by the combined occurrence of one or more anomalies. This is called a lead. By doing additional exploration, a lead may either be transformed into a prospect (developing the lead), or it may be wiped out.

Prospect

An anomaly or a combination of anomalies becomes a bonafide prospect when it meets a stated set of criteria considered requisite for a commercial accumulation of hydrocarbons. Once the presence or at least the potential presence of a source rock is established, there are two basic criteria that must be met:

the presence of a reservoir rock; and

the presence of a trap of sufficient size to hold a commercial quantity of producible hydrocarbons.

From the viewpoint of the explorationist whose job is to pick a specific drillsite, the presence of a trap is usually the more fundamental of the two, since it both locates and restricts the depth and areal position of the prospect. Also, it is usually the easiest factor to determine before a region has been drilled because potential traps can frequently be mapped through geological and geophysical surveys. However, many traps predicted by mapping have proved to be nonexistent after drilling. Poor well logs and samples, unsuspected facies changes, and faulty correlations all can lead to unreliable subsurface maps.

In making these maps, it's important to remember that prospects are located through creative geologic thinking and optimistic mapping. This is particularly true in mature petroleum provinces where others have been before. In mature regions with structural traps, structural contours can be redrawn to reflect possible additional closures ( Figure 1 and Figure 2 ),

Figure 1

Figure 2

or faults may be reoriented to suggest larger areal closures ( Figure 3 and Figure 4 ).

Figure 3

The chances of finding new fields with structural traps are sometimes slim in mature areas, however.

Figure 4

The exploration effort must then be directed toward finding subtle, stratigraphic traps. An example of such a trap is shown in Figure 5 and Figure 6 .

Figure 6

Figure 5

Here, porous oolitic limestones have accumulated in carbonate banks, forming porosity trends parallel to the paleoshoreline. Commercial quantities of natural gas occur within these porosity trends. Regionally, these banks follow NE-SW trends. In one area, though, the local paleodepositional trend is more northerly, as shown by the northeastern most porosity zone in Figure 5 . Recontouring on this basis extended the other local porosity zone farther to the north, and created an additional new prospect ( Figure 6 ). This was proved by subsequent drilling.

The second major concern in developing a prospect is the presence of a reservoir rock. We may know from past experience that a region contains just one potential reservoir. Other regions may have a production history that involves only a few out of a great many porous and permeable strata. Still other areas may be characterized by multiple stacked reservoirs. In a mature region, prospect generation may consist, to a large extent, of finding new traps in already well-established reservoir trends. We may not always be able to trace a good reservoir laterally, however. Facies changes, faulting, erosional disruption, or diagenetic changes can occur. These conditions are not always revealed by subsurface geological studies or by seismic methods.

We must emphasize, though, that a cluster of anomalies, however interesting, does not make a prospect unless the legal right to produce petroleum can be obtained either by lease, contract, or purchase.

False Anomalies

Some kinds of anomalies, particularly geophysical anomalies, may be misinterpreted. This can lead to false geological structures. "The primary error in conventional interpretation is the assumption that the seismic velocity is constant over a wide area. This simply is not so..." (Minturn 1982). If not recognized and corrected, local changes in the acoustic velocity of the rocks may lead to the generation of false structures. For example, a salt swell ( Figure 1 ) will give rise to a reflection time anomaly beneath the salt. Since acoustic velocity in salt is higher than in most clastic rocks, a false subsalt high is mapped.

Figure 1

In the opposite case, gas leakage and fracturing can cause abnormally low velocity. This can cause the crest of the structure to appear as a graben ( Figure 2 and Figure 3 ).

Figure 2

Velocity anomalies of this type cannot be ignored.

Figure 3

Either you use them positively to find productive fracture zones, or you compensate

for the erroneous structure which they induce. Similarly, magnetic anomalies that are due to differences in the magnetic susceptibility of various basement rock-types can be misinterpreted as being structures which have resulted from basement highs ( Figure 4 ,

Figure 4

produced by differences in magnetic susceptibility of basement rock types, and Figure 5 , produced by basement uplift).

Figure 5

Prospective Region or Trend

No oil company will begin exploring an area unless there is reason to believe it contains potentially commercial prospects that it is a prospective region or trend. Later, after the dust has settled, the company might be quite content to have found only one good field in the region. If one or two fields were all that seemed likely at the outset, however, exploration of that region might not have been worth the risk. Companies must play the odds.

A good discovery, X field in a fresh trend, or a new field from a deeper pay in an established region, inaugurates a new prospective trend, the X trend. The subsequent campaign of exploratory action may be called the X play. Large companies operating in United States provinces always aim to initiate new plays. Where competition exists, obviously, the major rewards of the exploration business come from receiving and applying specialized knowledge before competitors do, and particularly from acquiring relatively cheap leases in the prospective region early in the game.

In summary, then, prospect generation begins with the detection of one or more anomalies. When these anomalies indicate the possibility of a commercial deposit of hydrocarbons, a lead exists. This lead, in turn, may be developed into a bonafide prospect, or it may be discarded by conducting additional exploration. We must keep

in mind, however, that some anomalies, particularly geophysical anomalies, can easily be misinterpreted.

Assessing Prospective Regions

The crucial judgment for an exploration company is deciding whether an unexplored area is or is not a prospective region. If it is supposed to be prospective, but turns out not to be, much time and money will have been wasted. If the unexplored province is supposed to be non prospective, but it turns out to be prospective, the exploration company will have forfeited a chance for profit. A province is rarely written off, however, before some wildcats are drilled. Indeed, several dozen dry holes may be drilled before a province is called noncommercial.

Assessing prospectivity in producing regions is wholly different from that in frontier regions. In the former, you know you are in oil country, and the question is whether enough undrilled prospects remain between, beyond, above, and especially below known fields in order to justify further exploration. Much well information is usually available, and the main geological effort is geared toward answering local stratigraphic and structural questions before planning possible detailed geophysical surveys. On the other hand, in frontier provinces, the existence of the five essential factors (source rocks, reservoir rocks, migration paths, traps, and seals) for petroleum accumulation is a matter of speculation.

Many, if not most, of the world's remaining frontier provinces lie in offshore regions. Large untested provinces on land are apt to be in areas that are geographically or politically inaccessible.

While there are still some untested frontier provinces even in a mature producing region such as in the United States, several have disappeared from the list of potentially commercial provinces. The Lower Cook Inlet, Alaska, is one of these.

Data Needed

The most diagnostic exploratory tools in frontier provinces are wildcat wells and seismic surveys. On land, seismic is relatively expensive, and wildcat wells may be relatively less costly. In water-cove red regions, seismic is relatively less costly, but wildcat wells are very expensive. Photo geology is cheap but useful only over land. Aeromagnetic surveys are cheap and may be used either over land or sea to reveal basement structure and basin configurations. Gravimetric costs a little more and is a valuable prospecting and reconnaissance tool on land. Marine gravimeter surveys are inherently less useful because the magnitude of the corrections commonly exceeds the amplitudes of the anomalies.

Other tools are helpful in exploring certain provinces. These include:

side-looking radar, used like photo geology; particularly helpful in areas with heavy cloud cover and/or thick vegetation;

side-scan sonar, used for mapping underwater topographic-structural features;

Marine "sniffer" surveys, used to record and evaluate samples of hydrocarbon gases leaking from the sea floor;

soil hydrocarbon surveys, used to detect petroleum fluids leaking to the surface that produce halo patterns above buried reservoirs;

Radiation surveys, used to detect surface patterns of uranium daughter elements concentrated above buried reservoirs (Morse et al. 1982).

Technical Factors Assessing a frontier province begins with reconnaissance surveys on land, surface geology, coupled with photo-geology, is often used first. This is followed by aeromagnetic surveys, then gravimetric surveys, and finally seismic surveys. In marine provinces, after an initial aeromagnetic survey is made, a reconnaissance seismic survey is customary. The purpose of the geophysical reconnaissance survey is to uncover the kinds, depths, magnitudes, and relative frequency of buried anomalies in the province and to appraise the sedimentary sequence. It would be desirable at this point to make a preliminary inventory of indicated volumes and depths of potential traps. In practice, however, often the best one can do is to make semi quantitative generalizations and educated guesses about potentially productive closures and the total sediment thickness above basement.

One or more of the most obvious of these anomalies is evaluated by detailed seismic surveys. The most promising is then selected for the initial wildcat test. In most cases, pronounced structural anomalies are the first to be tested. In some offshore waters of the United States, "stratigraphic tests" are drilled deliberately off structure to directly observe stratigraphy, prior to federal lease sales. Usually this is a cooperative effort by several competing companies. Until these initial tests are drilled in representative locations, it may be difficult to discern the practical significance of certain geophysical anomalies.

Once drill cuttings, mud logs, and wire line logs are obtained from one or more wildcat wells, we can recognize the geological sequences present in the basin, detect hydrocarbon shows, and observe thickness and quality of reservoir strata. Moreover, it may be possible to determine the presence (or absence) of commercial source beds.

Model Prospective Region - Lower Cook Inlet, Alaska

The Lower Cook Inlet of Alaska provides a case study of an offshore prospective region. This province was thought to have commercial potential in the late 1970s. Figure 1 shows the situation at the time of the 1977 Federal Lease Sale.

Figure 1

In this case, the major job of geophysics was to map the shape, depth, and major structures of the province in order to compare the undrilled southern portion of the Cook Inlet basin with the productive northern portion. No attempt is made here to give either a history of the exploration of this province, or a full discussion of its geological problems. The only intention is to suggest how the exploratory prospects in the province were generated and what was learned from drilling.

Several marine seismic programs were undertaken during the 1960s and early 1970s. Most were sponsored jointly by a number of exploration companies. Syndication of risk is common in offshore provinces. Some exploration groups purchased two or more seismic surveys, plus gravimetric and aeromagnetic surveys. Each group sponsored an independent seismic interpretation. Because of the difficult multiple reflections and velocity problems encountered, supplementary side-scan sonar and shallow-penetration sparker seismic surveys were purchased from service companies.

The only published seismic maps of the Lower Cook Inlet were made by the U.S. Geological Survey, since all proprietary maps were kept confidential. Figure 2 displays the major positive structural features that were revealed, nearly all of which were investigated in much greater detail, revised, and confirmed by proprietary data.

Figure 2

Interested exploration companies decided which prospects were worth leasing and drilling partly by making an analogy to the Upper Cook Inlet. At long last, a federal lease sale was held. In October 1977, a total of 21 companies and groups of companies bid $398.5 million for 87 tracts (495,000 acres). By the end of 1979, dry holes were drilled on eight major prospects ( Figure 2 ). The Lower Cook Inlet province was found to be nonproductive.

One can imagine a frontier province being dry or noncommercial because of inadequacy or absence of reservoir strata, traps, or seals to the traps. By using reconnaissance seismic surveys, plus drilling a few wildcat wells, the existence of these three factors can be determined with considerable reliability. The presence or absence of these features are fairly straightforward determinations, but it is quite different with source beds, timing, and migration paths.

This point can be illustrated by a brief analysis of why prospects in the Lower Cook Inlet were dry. Although it is only a skeletal structure map, Figure 2 strongly suggests that large anticlinal closures (adequate traps) are common. Seismic cross sections, on which major reflection marker surfaces are identified stratigraphically ( Figure 3 , Seismic line 755, just north of Cape Douglas showing major reflection marker surfaces for stratigraphic correlation.

Figure 3

A horizon = base of Tertiary; B horizon = base of Lower (?) or Upper (?) Cretaceous; C horizon = Middle Jurassic (?) and minimum thickness of sedimentary rocks in basin), and structural sections ( Figure 4 ) both indicate that Tertiary rocks (Kenai Group) are present but thin in the Lower Cook Inlet.

Figure 4

Furthermore, the major oil pay of the Upper Cook Inlet (the Hemlock Conglomerate within the Kenai Group) seems to be present. It would appear then, that traps, seals, and reservoir rocks exist, and the abundant faulting, both old and new, which were shown by seismic lines, should have provided migration paths that were similar to those in the productive Upper Cook Inlet.

Evidently, source rocks that supply commercial amounts of petroleum are lacking in the Lower Cook Inlet. One obvious difference between the Upper and Lower Cook Inlet is that the lower Kenai Group, which is the principal objective, is very shallow (<1 km) south of the Augustine-Seldovia arch ( Figure 5 ), whereas it becomes progressively deeper towards the north.

Figure 5

The source beds of the Upper Cook Inlet oil fields are either the basal estuarine beds which floor the Tertiary sequence (Kelly 1968; Young et al. 1977), or the Middle Jurassic sequence which lies unconformably beneath the Tertiary (Claypool et al. 1980). Most oil companies seem to have accepted the idea of a Jurassic source, which was advocated by the U.S. Geological Survey. For either case, however, the sources in the Lower Cook Inlet are much shallower, hence much cooler and less mature, than in the north.

In assessing the shallow Tertiary anticlinal prospects in the Lower Cook Inlet sale area, most companies were not hopeful of substantial migration from Middle Jurassic source rocks as far upward as the Tertiary reservoirs. Tertiary seismic closures were regarded by most as secondary objectives at best. The Upper Jurassic clastic sequence was seen as a possible primary reservoir- if microfracturing due to recent uplift was adequate. The wildcat failures were evidence that microfracturing was inadequate for producing commercial-scale migration.

Estimating Hydrocarbon Potential and Calculating Risk

In Alaska’s Lower Cook Inlet, an area once thought to have great potential, the industry drilled eight dry holes. Overall perhaps a total of $750 million dollars was risked over more than a decade by at least 30 oil companies. It was all lost.

Assessing the risk and gross hydrocarbon potential of untested regions requires careful evaluation and complex analysis. A company can make a reliable estimate of how much money would be required to obtain a suitable position in a frontier province, but prior to taking that position, the company can't accurately forecast the expected rewards from the money risked. Substantial geophysical and geological reconnaissance, acreage acquisition, and wildcatting expenditures must be made before the region can be proved either commercially productive or barren. A major petroleum company must regard its share of the $750 million dollars spent in the Lower Cook Inlet as one of the costs of remaining in the exploration business.

The effective exploration technology developed in recent decades has shortened the time period needed to evaluate a frontier province. It has also made initial wildcat drilling highly successful where these new provinces have proved productive (in some cases, a 50% success ratio in wildcat drilling). A company would rarely go into a new province and spend such vast sums as in the Lower Cook Inlet, unless it had hopes of finding large, and possibly giant fields. The number of giant fields, even in major provinces, is relatively small. They are, however, most often found in obvious traps, such as anticlines, salt domes, or reefs. Often their location can be readily determined prior to drilling by geological or seismic surveys. Frontier provinces commonly provide the best chances for finding such first class prospects, though the risks and exploration costs are proportionately higher.

The Hepburn prospect (see the Example at the end of this section) is one such prime target. Its seismically mapped anticlinal closure with a "bright spot" assured it to be one of the first prospects drilled in the hypothetical frontier basin of the Southeast Spur. Prior to drilling, however, the findings of geological and geophysical surveys must be summarized in evaluation reports and prospect sheets. These present the fundamental geologic and economic facts concerning a prospect and also the initial estimates of anticipated reserves and economic returns. A sample prospect sheet for the Hepburn prospect is presented in Appendix A. On other prospect sheets, an attempt is made to assign grades or numerical ratings to the various factors in order to make the analysis more objective. Some formulas that are commonly used to calculate reserves, economic returns, and risk prior to drilling are presented below:

Reserves

Reserves = Oil-in-Place Percent Recovery

Oil-in-Place

Oil-in-Place (OIP) = Gross Reservoir Volume* Net/Gross Porosity Oil Saturation k** FVF**** Gross reservoir volume = productive areal closure reservoir thickness

** 7758 if acre-ft to bbl conversion

*** Formation volume factor

Potential Return (after tax) Future Net Revenue (FNR) = total reserves average price* - (operating costs + over total investment** + tax)* Over total payout time

** Cost of wildcat well, completion costs of discovery and any field development costs (sometimes this calculation excludes initial risk costs of wildcat well.)

Return on Risk Investment Return on Risk Investment = Potential Return Risk Investment*

* Predrilling cost + cost of wildcat well to casing point

Risk Adjusted Return on Risk Investment Risk Adjusted Return = (Potential Return / Risk Investment) Success Rate*

*based on past experience

Return on Investment (ROI) Return on Investment = Potential Return / Total Investment

In summary then, prospective regions that have seen little or no drilling may offer opportunities for some of the greatest rewards in finding and developing major new prospects and fields. The costs and risks, however, are significantly higher. Although potential traps, and reservoir and seal lithologies frequently can be detected through geophysical methods, often there are difficult fundamental questions to be answered. Some of these concern source rocks and their maturity, and the paths and timing of migration. These factors can make or break a potential prospective region. With appropriate planning, the technology is available to evaluate even a large frontier area rapidly, and either prove it to be an oil province or condemn it.

Example:

An Evaluation of Hepburn Prospect, Southeast Spur, Block 35, District #2

Petroleum Exploration Co. Exploration Dept. 5 Morrill Place Houston, TX 77036

Summary

i) Prospect is an anticline with four-way closure partially bounded by faults. Target is probably Jurassic sandstone.

ii) Seismic configuration is very promising. Five seismic lines pass through the prospect. The potential reservoir is a good reflector. The larger of the two culminations shows a "bright spot."

iii) Prospect is moderate to low risk with potential reserves of several ten million barrels.

Introduction This report describes the geologic and economic assessment of the Hepburn prospect carried out in November of the previous year. Five seismic lines delineate the prospect. Figure 1 (seismic line 1) shows a possible gas accumulation at 1.9 sec with a two-way time thickness of about 10 ms.

Figure 1

Figure 2 is a structural contour map on the top of the reservoir (Unit A) based on the seismic data.

Figure 2

Geology

Structure Low relief ESE-WNW trending anticline with two culminations. The structure is on a minor horst, bounded to the north and south by normal faults. Four-way closure, based on seismic data, has an estimated area of 4500 acres (18 km2) with 250 ft (76 m) of vertical closure.

Reservoir Unit A, a Middle to Upper Jurassic sequence of uniformly bedded alternating sand-shales throughout the 250 ft (76 m) of vertical closure. Unit thickness: 150 ft (46 m) Net/Gross Pay: Ranging from 0.3 to 0.7, most likely 0.5-0.6 Porosity: 12%-22%, most likely 17%

Seal Shales in the basal part of Unit B are required to provide a vertical seal for the prospect. Reflectors indicate good continuity in this unit. Assessment of laterally sealing faults is difficult.

Source Overlying Unit B shales. Prospect is well-situated for migration from these shales along down-to-basin faults.

Reserve Potential

Reservoir is estimated to be half-full to spill point (based on experience in Fowler basin): Productive areal closure* = 3050 acres (12 km2) Gross reservoir volume = 228,750 acre-ft (282 ´ 106m3)

Field Averages Minimum Estimate Maximum

Net/Gross ratio 0.3 0.6 0.7

Porosity 0.12 0.17 0.22

Oil saturation 0.6 0.65 0.7

FVF** 1.5 1.25 1.1

Net Reservoir Volume = Gross Reservoir Volume Net/Gross Ratio

Volume of Oil-in-Place = Net Reservoir Volume ´ Porosity ´ Oil Saturation ´ k***/FVF

*Measured from maps by means of planimeter.

**Formation Volume Factor is a factor that quantifies the change in volume of oil and gas when it is produced at the surface. It is mainly dependent on reservoir temperature and pressure.

*** k= 7758 bbl/acre-ft

Oil-in-Place

Minimum Estimate Maximum

106bbl 25.6 94.1 173.9

(106m3) (4.1) (15.0) (27.6)

Reserves (for estimated recovery of 28 %)

106 bbl 7.2 26.3 48.7(106m3) (1.1) (4.2) (7.7)

Economic Analysis

Recovery is estimated at 26.3 million bbl with limited water drive. Life of field = 30 yrs

Investment million $ Cost of 3000 m exploration well to casing point 1.95

(includes lease, geological, and geophysical costs) Exploration well completion costs 2.32 Total costs, discovery well 4.27 Field development costs (estimate 10 wells) plus surface equipment 50.00 Total Investment 54.27

Potential Return Potential gross revenue (based on 26.3 x 106 bbl @ $30/bbl) 789.00 Less: Operating Costs @ $2/bbl 52.60 Investment 54.27 Tax (50% net income after operating costs and investment) 341.07

Potential Future Net Revenue 341.07

Potential Return/Risk Investment 341.07 1.95 175/1

Risk Adjusted Return (based on 1.6 success rate) 175/11/6 29

Return on Investment 341.07 54.27 6.3/1

Play Analysis MethodsIntroduction

Ideally, a play analysis is the first step in planning an exploration program. We must investigate the individual play elements and present the information available about each element in an understandable format. Our goal is to identify areas within the basin where the play elements coexist in order to select prospective acreage.

We use play analysis to justify expenditures on exploration surveys, such as gravity and magnetic surveys, field investigations, and seismic acquisition. The play analysis continues until we have addressed all the risks associated with each play element. For example, suppose that we are investigating a new, deeper objective in a developed basin. This deeper section has been penetrated by only a few stratigraphic test wells, and we discover that the potential source rock within the section is overmature in all of these wells, or the potential reservoir units are heavily cemented. We will have serious difficulties persuading our management to spend more time and money continuing the play analysis. So, we must identify critical risk information early in a play analysis and present information about the play elements in an organized and timely manner.

We commonly use play summary charts, maps, and schematic play summary cross sections to show how the play elements in a basin relate to each other. To construct the summary, we use data from a variety of sources, including geologic maps, academic theses, journal articles, field sampling surveys, geophysical surveys, and subsurface well control. At the beginning of a play assessment in a new area it is essential that we collect all available data from previous investigations, and compile a preliminary play summary.

A basic play summary chart, which we may be able to fill-in early in analysis, might contain the following headings:

Age

Rock Units Lithology Thickness Source Seal Maturation Migration Trap Tectonic Setting Depositional Setting/System

Our preliminary play summary is particularly important in a frontier area because it is a critical planning tool that can show us what data we have and what data we are missing.

The play summary chart cannot show detailed information about a given play element. While it can indicate that a Cretaceous reservoir with 20 percent porosity exists in the basin, only a map can show the areal distribution of the reservoir unit, its thickness, and variations in porosity and permeability. A summary chart can indicate Tertiary thrust faults as a potential structural trap element, but we need a map to show the distribution of potential structures, their area, and estimated closures. In a frontier play, many of these maps will be impossible to construct because of sparse data; again, our preliminary set of maps will serve as a planning tool. Finally, we can use schematic, play-summary cross sections to show relationships between play elements at depth.

Sources of Information

There are numerous sources of information that we should incorporate into our play analysis. These include surface geological and geochemical data, seismic, gravity and magnetic surveys, as well as well control. Let’s consider each of these data sources in turn.

Surface geological control includes maps (geologic, soil, hydrologic, topographic), aerial photography, satellite imagery, measured sections, outcrop samples, and cross sections generated from surface measurements. We can determine lithologic characteristics, including rock type, porosity, permeability and diagenesis, from measured sections, outcrop samples, and geologic maps. These data sets also contain direct and implied information on depositional environments and facies relationships in reservoir, source, and seal units. We can measure organic matter content and type, as well as maturation, in outcrop samples of source units, and paleontological age indicators measured from lithological samples. Geologic maps also contain information on the areal distribution and thickness of stratigraphic units, and measured sections show areal variations in thickness.

Geologic maps contain the structural information we need to understand trap geometries. The maps may show exposed structures with measured strikes and dips, or surface expression may be subtle where undeformed sediments overlie structures. In the latter case, geomorphic expression may be a guide to buried structures, so topographic maps can be a valuable exploration tool. We can combine information from geologic maps and measured sections with topographic maps to construct cross sections showing stratigraphic and structural relationships at depth. These surface data sets form the foundation for the interpretation of subsurface data, and are critical to play analysis in frontier regions where subsurface data may be scarce.

We can divide surface geochemical sampling surveys into two categories -- the study of source rocks from samples, and the detection of hydrocarbon seepage at the surface.

In frontier play analyses, the primary goal of source rock studies is to determine the presence of a source rock capable of generating petroleum. This initially involves analysis of three parameters: organic matter quantity, organic matter quality, and maturity. In frontier areas, surface sampling may be the only way to access potential source rocks even if well control exists, because samples from existing wells may not be available for analysis. We should always select samples with minimal surface weathering effects. Surface source rock samples are first put through a screening process called whole-rock pyrolysis, which provides information on organic matter quantity, type, and estimated maturity. We can confirm the maturation level and organic matter type using visual characterization of the kerogen separated from the samples, since organic matter darkens as maturation increases.

A geochemical report of source rock analysis will probably contain many pages of data summaries. Do not be overwhelmed. For a quick reference, Table 1 lists these source rock pyrolysis parameters.

S1Measure of bitumen (matured kerogen) in source rock sample. Increases with increasing maturity.

S2Measure of kerogen remaining in source rock sample. Decreases with increasing maturity.

S3Measure of carbon dioxide generated from kerogen during pyrolysis. Related to oxygen content in kerogen.

TmaxTemperature at which S2 generation is highest. Increases with increasing maturity.

%TOC Percentage of organic carbon in source rock sample.

HI Hydrogen Index. S2/TOC is a measure of unrealized hydrocarbon generative potential in source rock sample.

OIOxygen Index. S3/TOC is an approximation of oxygen content in kerogen. OI can be plotted against HI to analyze kerogen type and maturation level.

Table 1: Key source rock pyrolysis parameters.

When performing surface geochemical prospecting, our goal to detect microseepages of oil and gas. There is general agreement that many petroleum accumulations leak hydrocarbons to the surface via faults, fractures, unconformities, intrusions, and highly permeable sediments (Hunt, 1981). Price (1986) asserts that C1 to C5 hydrocarbons migrate to the earth’s surface from thermogenic hydrocarbon deposits to form anomalously high, measurable surface concentrations, and that C2 to C5 hydrocarbons are uniquely indicative of thermogenic hydrocarbon deposits. Price favors vertical migration of microbubbles of natural gas through microfracture

systems which overlie hydrocarbon deposits as the transport mechanism for microseepages.

As they migrate to the surface, hydrocarbons impose chemical, biologic, and mineralogic effects, known as geochemical anomalies, on the overlying rock and soil column. We can detect these anomalies by direct geochemical prospecting techniques that measure relative or absolute C2 to C5 concentrations in the surface and near-surface environment. These techniques include adsorbed and occluded soil-gas analysis, free soil-gas analysis, microbial analysis, and integrative absorption analysis. Table 2 briefly summarizes these techniques. Indirect geochemical prospecting techniques, which detect physical and chemical changes in surface and near-surface soil caused by vertically migrating hydrocarbons, include radiometric, geobotanical, and magnetic surveys ( Table 3 ).

Soil-gas Hydrocarbon analysis:

Measure of C2-C5 hydrocarbons in soil-gas samples.

Soil-sorbed Hydrocarbon analysis:

Measure of hydrocarbons absorbed by soil particles such as clays.

Soil-occluded Hydrocarbon analysis: Measure of hydrocarbons occluded in soil carbonates.

Microbiologic analysis:

Measure of microbial activity related to microbial consumption of hydrocarbons.

Integrative absorption:

Use of absorbents (i.e. charcoal) to collect hydrocarbon gas flux through soil over an extended period of time (days to weeks).

Table 2: Direct geochemical detection methods.

Radiometric survey:

Measures a radiation "low" over some oil reservoirs, possibly caused by absorption of radioactive materials in reservoired oils.

Geobotanical survey:

Detects changes in vegetation type, vigor, and density, porosity caused by soil-gas anomalies that may create soil reducing conditions over reservoirs.

Magnetic survey:

Measures increased magnetism over some oil reservoirs, possibly caused by reducing soil conditions from microseepage, which may reduce non-magnetic hematite in sediments and soils to magnetic hematite.

Table 3: Indirect geochemical detection methods.

Surface geochemical prospecting surveys have been applied to hydrocarbon exploration for over fifty years, yet they are still controversial. Hunt (1986) states that geochemical anomalies are best used in a regional sense since there is no known mechanism that will cause a subsurface pool to be outlined at the surface. Regional geochemical surveys can be risky but can indicate whether a basin has generated hydrocarbons. There are numerous published examples of successful applications of geochemical prospecting surveys (i.e., Bond, 1988, Jones and Drozd, 1983), but it is important to recognize the many limitations of geochemical prospecting.

Surface geological and geochemical surveys and source rock studies are key ingredients in a frontier play analysis. These data form the foundation for interpretation and planning of other surveys that investigate the subsurface -- geophysical surveys and drilling.

We use magnetic surveys in frontier play analysis primarily to determine the thickness of the sedimentary section within a basin. We also use these surveys to locate intrusives and buried volcanics, as well as to detect faults and other structures. Magnetic surveys are typically airborne and may cover thousands of square miles. These surveys enable us to economically identify basins or parts of basins with adequate sediment thickness whose hydrocarbon prospectivity we may want to assess further. Nettleton (1971) presents a representative application of a regional magnetic survey in Senegal. The geologic map of Senegal ( Figure 1 ) suggests a basinal area covered by alluvium and Pliocene deposits, between outcrops of Precambrian basement near Dakar and on the southeast.

Figure 1

A basement depth map ( Figure 2 ) calculated from a reconnaissance aeromagnetic survey indicates two basins with deep sections (>7000 m) on the southwest and northwest, a shallower basin on the southeast, and a northeast-trending platform between the basins.

Figure 2

This survey provides a quantitative picture of the gross features of the basin from which we can begin an evaluation of its petroleum prospects. Individual magnetic highs and lows may correspond to basement relief, but they may also represent contrasts in magnetization within basement or sedimentary rock units. In order to use magnetic data for detailed studies, we must interpret these anomalies and integrate them with other data sets.

Gravity surveys have a similar application in frontier play analysis -- namely to outline regional basin trends and estimate depths. Gravity data are a more reliable indicator of basement relief, and of structural configuration, than are magnetic data. Warsi (1990) defined the regional structural framework of Kuwait based on a regional gravity survey. The geologic map of Kuwait ( Figure 3 , Generalized geologic map of Kuwait.

Figure 3

1=Holocene-Quaternary cover comprising eolian sand, alluvium and coastal deposits; 2=Dibdibba Formation [pebbly sand, gravel and poorly sorted sandstone]. 3=Fars and Ghar formations [sand, subordinate clay, and nodular limestone]; 4=Damman Formation [limestone with chert]. The Damman Formation crops out very locally along the Ahmadi Ridge.) indicates extensive unconsolidated cover (units 1 and 2) and a limited surface expression of subsurface structures. The Bouguer gravity anomaly map ( Figure 4 ) defines two major arches, the Kuwait Arch and the Dibdibba Arch, and an intervening basin, the Dibdibba Basin.

Figure 4

These structures reflect basement relief, based on magnetic data and well control. Both the Kuwait Arch and the Dibdibba Basin are productive; further exploration is probably warranted along the Dibdibba Arch, which is similar in structure and stratigraphy to other productive arches in eastern Arabia (Warsi, 1990).

We may not be able to uniquely identify many anomalies we interpret from separate gravity and magnetic surveys, because both gravity and magnetic anomalies can be generated from a variety of subsurface sources that produce similar signatures. By combining the two data sets, we can often resolve ambiguous data. For example, Kulik (1990) found a major gravity low that correlates with the surface distribution of low-density Absaroka volcanic rocks in northwestern Wyoming. Magnetic values over the northern part of the field show characteristic high-amplitude anomalies, indicating thick volcanics. Magnetic values over the southern area are flat, indicating thin volcanics. This suggests that the gravity low in the southern area correlates with thick sediments in a basin beneath thin Absaroka volcanic cover.

We can also apply magnetic and gravity data to the regional analysis of thrust belts. Hartman et al. (1982) integrated magnetic and gravity data to determine the depth and relief of thrust sheets in the Appalachian overthrust area. They were also able to map the relief on the basement surface, and to identify thrust sheets containing wedges of crystalline basement. Rasmussen (1985) recommends magnetic survey interpretation in areas where overthrusts contain magnetic basement rocks. He

determined the thickness of basement thrusts overlying sediments, as well as the dip of the base of the overthrust along the Arbuckle Mountain thrust front in Oklahoma. Rasmussen concludes that, by using magnetics data to detect and map igneous overthrusts, we can find undrilled sedimentary areas large enough to contain prospects of significant size and reserves.

Gravity and magnetic models of geologic cross sections are powerful tools in regional geologic interpretation. Silver et al. (1982) interpreted gravity data of the northern Idaho-Wyoming Overthrust Belt, constructing gravity models to constrain the structural configuration of basement and overlying sedimentary veneer on geologic cross sections based on surface data. Their integration of surface geologic and gravity data differentiated between pre- and post-thrust basement uplifts that influence subsurface structural geometry and stratigraphic thickness. This type of data integration, where modeling constrains subsurface interpretation, is vital in frontier areas where subsurface control from seismic and wells is scarce or nonexistent. Chandler et al. (1989) modeled both gravity and magnetic surveys with sparse seismic data to develop a regional tectonic framework for application to petroleum exploration in the Mid-Continent Rift System of North America. By incorporating constraints from gravity and magnetic data, they were able to differentiate between parts of the rift with predominantly volcanic fill and those with predominantly sedimentary fill.

If available, seismic data can provide a valuable source of subsurface control in a frontier play analysis. If seismic data of any vintage exist, do not ignore them. Do not pass over older data or poorly processed data; these can often be reprocessed into a useful format. Even poor seismic data may show geometric patterns that delineate structural and stratigraphic relationships. In frontier areas, any subsurface information about play elements is critical.

Seismic surveys have been used for decades to identify structural and stratigraphic traps. Figure 5 (A salt wall diapir [D] from the North Sea with some of the possible trap positions shown.) is a classic example of potential trap delineation in an area of salt tectonics.

Figure 5

Although this type of salt structure would have a recognizable signature on geologic and topographic maps, and gravity and magnetic surveys, the structural detail necessary to identify various traps is available only on seismic data. Similarly, the structures in Figure 6 (Uninterpreted and interpreted seismic sections through the South Elk Basin producing area [northeast Big Horn Basin,

Figure 6

Wyoming], showing anticlinal folding over basement thrusts.) would have surface expression and gravity/magnetic signatures, but we can interpret the dip of the faults and trap geometry in greater detail using seismic data.

Seismic interpretation has moved beyond simple trap identification and into the identification of structural styles. This application is particularly important in frontier play analysis. Regional seismic coverage, if properly oriented, can be critical in structural style interpretation. Figure 7 (Uninterpreted and interpreted seismic sections from offshore western Africa, showing a complex diversity of both erosional

and structural terminations in reflection patterns.

Figure 7

Letters indicate ages of units: TPL=Tertiary, Pliocene; TM=Miocene; TP&E=Paleocene and Eocene; K=Cretaceous; J=Jurassic; TR=Triassic.) and Figure 8 show a succession of structural styles in a section from offshore western Africa.

Figure 8

The Triassic-Jurassic units are affected by basement-involved extensional faulting, while the Cretaceous sequence is affected by detached extensional faulting. The Tertiary section is unfaulted. This sequence of structural styles suggests rift formation followed by the development of a passive margin tectonic habitat. Interpretation of regional structural style and tectonic habitat aids in basin classification as well as in better prediction of potential trap distribution and geometry.

We can also use regional seismic coverage to interpret depositional systems. Figure 9 (Seismic reflections [top] and geological interpretation [bottom] of Upper Cretaceous carbonate shelf margin,

Figure 9

Aquitane Basin, offshore France.) illustrates the seismic expression of a carbonate shelf margin, and the corresponding geological interpretation based on analogous depositional systems. Bubb and Hatleid (1977) summarized the seismic criteria for recognizing carbonate depositional systems. Similar criteria have been developed for the seismic recognition of prograding sequences on clastic shelves ( Figure 10 ), and for clastic sediments prograding onto the basin plain.

Figure 10

Once we have recognized a depositional system, we can use seismic stratigraphy to predict where reservoir, source, and seal units occur. One of the fundamental precepts of sequence stratigraphy is that a seismic reflection represents a time plane, and that patterns of reflection time show depositional sequences. A sequence is bounded by unconformities and their correlative conformities. Sequences are divided into systems tracts, which are deposited at different sea levels related to eustatic sea level changes. Each systems tract comprises a group of contemporaneous depositional systems. For example, Figure 11 shows the lowstand systems tract, which includes marginal marine to basin plain depositional systems, subdivided according to rock type.

Figure 11

Figure 12 shows a summary of the reservoir potential of the lowstands systems tract.

Figure 12

Figure 13 shows the reflection patterns that comprise a complex shelfal system,

Figure 13

and Figure 14 (a) shows these stratal patterns interpreted in terms of likely rock type.

Figure 14

Figure 15 (b) also shows a rock type/stratal pattern diagram redisplayed with geologic time as the vertical scale.

Figure 15

This chronostratigraphic chart is a valuable record of the geologic history of a prospective basin, particularly if it can be calibrated to real geologic time. With adequate paleontologic control this may be reasonably easy, but in a frontier basin it may require the generation of sea level curves showing the variation in eustatic sea level related to the different systems tracts. These curves can be matched to similar

curves generated in dated sediments in other basins. Anstey (1992) concludes that under favorable circumstances, we can use this approach to obtain the age of deeply-buried rocks from the patterns of their seismic reflections.

We can use seismic sections to evaluate trap geometry and the distribution of reservoir, source, and seal, and, in certain cases, to evaluate maturation. Figure 16 , a seismic section from a complex area of salt tectonics, has been used by Anstey (1992) to illustrate a geologic history analysis.

Figure 16

He begins with a series of historic reconstructions of the various stages of tectonics and sedimentation, which he uses to construct a burial-history diagram ( Figure 17 , The burial history diagram for the trough region between the salt pillows of Figure 13, constructed from stage diagrams.

Figure 17

The depths are obtained from the seismic time intervals using reasonable velocities, and with due regard for compaction in the shales. The geologic time periods are in millions of years before present.). Based on the burial history, Anstey concludes that Tertiary (T) source rocks are immature; Lower Jurassic source rocks (J) reached maturity (depths below 1500 m) about 30 My ago; and pre-Zechstein (pre-Ze) source rocks have been mature for over 200 My. This analysis provides a measure of the timing of maturation and the timing of structural trap formation. By determining the possibilities for migration pathways, we can also evaluate the potential for hydrocarbon accumulation.

Two types of data are available from wells. We obtain direct information about reservoir, source and seal units, maturation indicators, and hydrocarbons from rock and fluid samples from cores and cuttings. Indirect measures of rock and fluid parameters come from wireline logs. In a frontier basin, well control may be minimal or non-existent, but if available it can provide key subsurface control for lithologic and paleontologic analysis, geochemistry, correlation, and interpretation of seismic and other geophysical surveys.

Rock samples from wells come in three forms: cuttings, sidewall cores, and conventional cores. Cuttings are collected during drilling and used to prepare sample logs, based on visual examination. Sample logs provide a written lithologic and paleontologic description of the rock units penetrated and note any oil shows

present. These logs can be used for environmental analysis, dating, and correlation. If cuttings themselves are available, and have been stored properly, screening analyses can be run on potential source units. We can use sidewall cores for similar analyses. Conventional cores provide continuous samples over the cored interval and afford the best samples for lithologic and paleontologic analysis. They often show sedimentary structures which we can interpret for environmental study and paleocurrent information. We can also make porosity and permeability measurements using conventional cores.

Wireline logs measure physical parameters such as resistivity, density, velocity, spontaneous potential, and radioactivity. Log interpretation translates these parameters into measures of permeability, porosity, hydrocarbon saturations, and lithology. We use these measures to identify potential reservoir and seal units. Subsurface lithologic sections interpreted from well logs can be correlated with surface sections, as well as with sections in other wells. We use correlated sections to construct cross sections showing structural and stratigraphic relationships. We then use these sections for environmental analysis to evaluate depositional settings, systems and facies, and to predict the distribution of reservoir, source, and seal units. We can also use these sections for structural analysis, and to interpret trap geometry and distribution. We can construct geologic history analyses from well log cross sections in a manner similar to that described for seismic data. Other important applications include structure contour mapping and isopach mapping.

Well data also provide valuable controls for geophysical surveys. Samples can provide density and magnetization measures to aid in interpretation of gravity and magnetics surveys. Wells drilled to basement can be particularly useful for this purpose. Density and sonic logs provide direct measures of density and acoustic velocity, which we can use to generate synthetic seismograms. The synthetic seismogram provides direct correlation between rock units and seismic sections, aids in time-to-depth conversions of seismic data, and in seismic correlation and modeling.

In frontier play analysis, we typically rely heavily on analogs to help predict geological conditions in poorly explored basins. Productive basins in close proximity to a new area are the most attractive analogs. We have cited examples of production from a variety of different types of basins, associated with several structural styles and depositional systems. We can use these and other producing examples to predict potential plays and to model the distribution of play elements in our area of interest. Selecting an appropriate producing analog, based on play controls, may be critical to establishing the credibility of our play analysis.

Play Summary Development

The first step in a play summary is to organize and evaluate the data available for each play element. If we already have a range of geological, geophysical and well data available, we are probably trying to generate new plays in a previously-explored basin. However, in a frontier basin we will almost certainly need to obtain new data to fully evaluate plays. A preliminary play summary chart), which summarizes all available data, lays the foundation for planning data acquisition, upgrading the play analysis, and identifying leads. We will look at the geologic and gravity maps of Kuwait ( Figure 3 (Generalized geologic map of Kuwait.

Figure 3

1=Holocene-Quaternary cover comprising eolian sand, alluvium and coastal deposits; 2=Dibdibba Formation [pebbly sand, gravel and poorly sorted sandstone]. 3=Fars and Ghar formations [sand, subordinate clay, and nodular limestone]; 4=Damman Formation [limestone with chert]. The Damman Formation crops out very locally along the Ahmadi Ridge.) and Figure 4 ) to create a hypothetical example of a frontier play analysis.

Figure 4

A comprehensive play summary takes into account every important aspect of the source, migration, reservoir, trap, timing, and seal play elements (White, 1988). Table 1 lists the variety of play summary maps produced during a comprehensive exploration mapping program. This level of mapping indicates a thorough understanding of the various play elements, which can be achieved with adequate planning and data acquisition. Even if we begin our play analysis with very little concrete data about most of the play elements, we must address these elements as our play analysis evolves. For example, our preliminary play summary might indicate oil seeps within a topographic basin whose trend parallels nearby productive basins. In such a situation, the productive analogs may provide predictive data on key play elements, such as potential source, seal, reservoir, and trap type. We might use this level of play summary to justify lease acquisition, and as a basis for planning data acquisition. We then integrate these new data into new play summary maps, charts, and cross sections, which, in turn, guide further data acquisition and analysis.

HC-Control Factor Examples of Possible Maps

SourceBed thickness and Effective isolith (thickness drained)

area Total organic carbon Organic matter type Maturation Overmaturation Combination of above

TOC percent to cutoff Adequacy edge: oil-vs. gas-prone Windows form maturity indicators HC deadline HC volumetric yields

MigrationSecondary migration (primary under Source)

Paths from structure, stratigraphy Prospective perimeters from source

ReservoirGross thickness Net/gross Porosity Permeability

Isolith with effective edge Net/gross ratio to cutoff Porosity percent to cutoff Appropriate facies types

Trap and TimingClosure area, height Timing

Structural contours, strat pinch-outs Trap timing keyed to migration timing

SealThickness Lithology Modifiers

Isolith with effective edge Facies type, entry pressure, ductility Faults, fractures, hydrodynamics, tar sealing, diagenesis; gas hydrates

Preservation and RecoveryFlushing Biodegradation Diffusion Viscous oil Inert-gas dilution Insufficient Concentration

Hydrodynamics; salinity; solubility Formation-water types HC type, seal diffusivity, timing Oil viscosity or gravity Inert fraction of gases Bbl/acre or bbl/platform, by trap

CombinationHydrocarbon occurrence Fields, shows, seeps, seismic signs

Tested and untested closures Illustrative cross sections Play summary map: Key boundaries, favorable areas, success ratios

Table 1: Play elements and related maps.

A typical frontier play analysis begins with the recognition of a prospective basin. Data search and compilation are the first steps in the analysis. Photogeologic

mapping at a scale of 1:500,000 to 1:250,000, including field checking and sample collection, is a practical next step. A regional aeromagnetic survey, followed by a focused gravity survey, will indicate basin configuration, sediment fill, and structural trends. Regional seismic lines should cross the basin axis and margins, followed by in-fill seismic coverage that concentrates on areas where potential plays are best developed. We select these areas based on the integration of surface and subsurface data sets. We modify and adapt the seismic program as new data are acquired. This approach has proven to be the most cost-effective way to concentrate high-cost exploration methods (seismic, drilling) in the most prospective areas of a basin. Our play summary evolves with the acquisition of each new data set, and gradually grows into a comprehensive set of maps and cross sections that we can use to generate leads and drillable prospects.

There are exploration environments in which this sequence of surveys is not possible. Surface geologic maps are not an option offshore, for example, but bathymetric maps and side-scanning sonar may indicate the expression of structures beneath the sea floor. Seismic data may be difficult or impossible to obtain under certain surface conditions, such as intense karstification or extensive sand dune or volcanic cover. In such circumstances, a careful integration of all data sets is critical for play analysis.

We begin our play analysis of Kuwait with two basic data sets that cover the area -- a geologic map ( Figure 3 ) and a Bouguer gravity map ( Figure 4 ). A few magnetic traverses are also available. The geologic map is not particularly enlightening because of extensive Pleistocene-Holocene unconsolidated sedimentary cover. The Bouguer map is encouraging, however, because the north-south-trending Kuwait arch is analogous to highly productive structural arches in Saudi Arabia, Iran, and Iraq (Warsi, 1990) The northwest-trending Dibdibba Arch is confirmed as another basement uplift based on its integrated gravity/magnetic signature. The intervening Dibdibba Basin is identified by a north-northwest-trending gravity low, which includes several small closures that may indicate basin deeps. Our preliminary play summary, while admittedly sketchy, indicates a favorable structural configuration for traps, and potential deep basinal areas suitable for maturation. We still need information on the elements of source, seal, and migration, and more details on the trap and maturation elements.

In this example, we can address many of our questions about source, seal and reservoir by analogy, because these elements are favorably developed regionally in highly productive trends in Iran, Iraq, and Saudi Arabia. However, the analogous lithologies are not exposed at the surface in Kuwait, and we require subsurface confirmation of their distribution for depositional system analysis. We also require subsurface detail to confirm the configuration of the basement uplifts and basin we mapped on the gravity data. Figure 3 shows a reconnaissance seismic survey (600 km) designed to evaluate the structural configuration interpreted from gravity data. This survey will also reveal subsurface stratigraphic relationships, which we can compare with the regional seismic and well log coverage from the surrounding region to model and interpret the subsurface section. At this point we should have good control for the play analysis, correlated to productive regional analogs. We can use in-fill seismic data to evaluate any favorable leads identified on the reconnaissance seismic survey, and possibly bring these leads to the prospect stage.

If we must begin a play analysis in an area where there is no nearby production, no seismic data, and no well control, our play analysis will require different planning procedures. For example, if we begin with the geologic and magnetic maps of

Senegal ( Figure 1 and Figure 2 ),

Figure 1

the same play elements will be much more difficult to analyze because there are no productive basins nearby.

Figure 2

Here, the magnetic survey suggests that three basinal areas are separated by an interior platform. Nettleton (1971) compares the structural configuration of Senegal to that of the highly productive Central Basin Platform and adjacent Delaware, Midland, and Valverde basins in western Texas. This analogy is attractive, but must be developed with more data. The first task in an area like this should be a new surface geologic interpretation. Geologic maps of frontier areas are often inaccurate. They usually fail to show any geomorphic expression of structures, and often fail to show numerous outcrops present along drainages. Attempts to plan field sampling and mapping programs from small-scale (1:500,000 or smaller) maps are likely to miss important locations for data gathering. Our interpretation of stratigraphic and structural relationships may also suffer if we rely exclusively on available maps. A regional photogeologic interpretation from high-resolution satellite imagery, such as Landsat, SPOT or ERS, will provide a more accurate geologic base map for planning necessary field, magnetic, gravity, and seismic surveys, and for interpreting and integrating data from these surveys.

We should plan surface geological and geochemical surveys based on accurate geological maps. Surface outcrop sampling is essential when evaluating potential source, reservoir, and seal units. Areas of thin alluvial cover may be candidates for shallow stratigraphic core hole drilling. Surface expression of structures should be

field checked, and surface fault/fracture zones may be candidates for direct geochemical prospecting surveys.

We will need more detailed subsurface information to completely evaluate structural and stratigraphic elements. We have outlined several potential basins on our magnetics data, but seismic coverage would be too expensive to cover all of the potential areas of interest. We can use a regional gravity survey to confirm basin trends and depths, and to indicate major structural features within the basins. We can use this gravity survey to focus our future seismic efforts in areas with the highest potential. We then plan a series of regional seismic lines across these areas to investigate structural styles and depositional systems, to predict potential trap configurations, and to project the distribution of potential reservoir, source and seal units. Several plays may become apparent at this point. Integrating the various subsurface and surface data sets will further narrow the focus of exploration into the area or areas with the most prospective leads. We will then cover these areas with more detailed seismic surveys and use the results of this more detailed survey to develop prospects.

Play Summary Presentation

There are only a few published examples of play summaries. While many articles address one or two play elements, few present an integrated play summary. An analysis of the hydrocarbon potential of the Bass Basin, Australia done by Williamson et al. (1987) provides a good example of the kinds of maps and cross sections that comprise a play analysis. Their study, which is based on regional seismic coverage and limited drilling, includes structural leads maps for several prospective reservoir horizons ( Figure 1 (Structure map for [?] Jurassic pre-rift unconformity level.) and Figure 2 (Structure map for Paleocene [top.

Figure 1

L.

Figure 2

balmei zone] level of Eastern View Coal Measures.), and maturation maps for two potential source rocks ( Figure 3 (Maturity map for base of Eastern View Coal Measures at 40 Ma.)

Figure 3

and Figure 4 (Maturity map for top of Eastern View Coal Measures at 40 Ma.

Figure 4

). They also illustrate the evolution of depositional systems in the Bass Basin with a series of block diagrams ( Figure 5 and Figure 6 , Schematic evolution of sedimentary environments of the Bass Basin.

Figure 5

a] Middle Cretaceous base Eastern View Coal Measures [EVCM].

Figure 6

Immediately postdates formation of basin by extensional tectonics. Alluvial-fan and braided-stream sedimentation is dominant; part of graben probably drained internally and formed lakes. Sedimentation rates were high, more than 50 m/my [160 ft/my] . b] Paleocene L. balmei horizon of EVCM. Most basement highs drowned, and extensive flood plain with marginal alluvial fans developed. c] Late Eocene [Demons Bluff Formation]. An extensive low-gradient flood plain passed westward into tidal-reach environment during rapid marine transgression. d] Ogliocene [Torquay Group]. Establishment of shallow marine conditions led to carbonate shelf deposition, which persists to present day.). Other illustrations show distribution of source rocks, geohistory curves, and representative seismic and well log sections. Although there is no final play summary map, the maps, cross sections, and diagrams effectively address all of the play elements.

Wanli (1985) also addresses all of the play elements in his study of the Songliao Basin in China. He maps structural subdivisions within the basin ( Figure 7 , Structural location map of Daqing oil field within principal depression of Songliao Basin.

Figure 7

Legend: a] boundary of principal depression of Songliao Basin; b] boundary of central basin high; c] oil-bearing area [numbered pools are identified in upper left corner]; d] gas-bearing area; e] depression; f] terrace. Inset: location map of Songliao Basin and surrounding structures; 1] slope from the west; 2] plunge from the north; 3] central basinal area; 4] uplift on the northeast; 5] uplift on the southeast; 6] uplift on the southwest.), as well as source and reservoir distribution, maturation levels of several source units, hydrocarbon productivity, and the relationship between maturation and structural development (timing) in the basin

( Figure 8 and Figure 9 ,

Figure 8

Relationship between area of main mature source rocks at different geologic times and development of the Daqing high.

Figure 9

Legend: 1] maturation area of source rocks of Member Qing 1; 2] maturation area of source rocks of Members Qing 2 and 3; 3] maturation area of source rocks of Member Nen 1; 4] top of Members Yao 2 and 3; 5] present boundary of oil-bearing area.) The key play elements are also presented in a regional cross section ( Figure 10 , Cross section of Zhenlai to Shengping area showing source, carrier, and reservoir rocks [middle oil-bearing assemblage] in Songlaio Basin.

Figure 10

Legend: 1] source rocks; 2] reservoir rocks; 3] cap rock; 4] oil pool; 5] direction of oil and gas migration; 6] direction of subsurface water flow.). His final play summary map and cross section ( Figure 11 , Map showing coordination among five bodies in Gulong-Daqing-Sanzhao area.

Figure 11

Legend: 1] area of source rocks; 2] boundary of central basin high; 3] 75-m isopach for sandstones in Sa, Pu, and Gao oil reservoirs; 4] cap rock; 5] axis of depression; 6] oil and gas carrier bed and reservoir; 7] 5-m contour for sandstone in Putaohua oil reservoir; 8] huge sand body; 9] extent of Daqing oil field.) shows the relationships between the play elements as they exist in the giant Daqing oil field.

White (1988) demonstrates the applications of oil and gas play maps and cross sections to exploration. He lists the play elements that we must study in order to analyze a play, and the maps that can be used to illustrate the different elements ( Table 1 ).

Figure 12

Figure 12 and Figure 13 (Mission Canyon limestone plays on anticline, noses, subunconformity closures, and fracture traps.

Figure 13

Williston Basin, United States and Canada.) shows a regional play summary map and cross section compiled by White for several Mission Canyon limestone plays in the Williston Basin. It integrates all available data concerning source, migration, reservoir, trap, timing, and seal, and subdivides the area according to the quality of the various elements. This kind of regional play map is the optimal tool for indicating the best areas to explore, and should be the goal for summarizing a frontier play analysis.

HC-Control Factor Examples of Possible Maps

SourceBed thickness and area Total organic carbon Organic matter type Maturation Overmaturation Combination of above

Effective isolith (thickness drained) TOC percent to cutoff Adequacy edge: oil-vs. gas-prone Windows form maturity indicators HC deadline HC volumetric yields

MigrationSecondary migration (primary under Source)

Paths from structure, stratigraphy Prospective perimeters from source

ReservoirGross thickness Net/gross Porosity Permeability

Isolith with effective edge Net/gross ratio to cutoff Porosity percent to cutoff Appropriate facies types

Trap and TimingClosure area, height Timing

Structural contours, strat pinch-outs Trap timing keyed to migration timing

SealThickness Lithology Modifiers

Isolith with effective edge Facies type, entry pressure, ductility Faults, fractures, hydrodynamics, tar sealing, diagenesis; gas hydrates

Preservation and RecoveryFlushing Biodegradation Diffusion Viscous oil Inert-gas dilution Insufficient Concentration

Hydrodynamics; salinity; solubility Formation-water types HC type, seal diffusivity, timing Oil viscosity or gravity Inert fraction of gases Bbl/acre or bbl/platform, by trap

CombinationHydrocarbon occurrence Fields, shows, seeps, seismic signs

Tested and untested closures Illustrative cross sections Play summary map: Key boundaries, favorable areas, success ratios

Table 1: Play elements and related maps.

Sudan Rift Basin

Seven years of carefully-planned and focused exploration resulted in Chevron’s 1980 oil discovery, Unity #2, in the Muglad Basin of southern Sudan. Chevron’s initial interest in southern Sudan was sparked in February, 1973, based on a regional photo geologic interpretation of satellite imagery designed to map structural trends in Kenya and surrounding areas. The interpretation suggested a structural trough or graben in the southern part of Sudan. This potential new basin in an unexplored region prompted a data-gathering trip to Sudan, which yielded a small aeromagnetic survey carried out by the United Nations as part of a groundwater study. The aeromagnetic survey indicated that a graben was present, and that it probably

contained up to 3000 meters of sediment. Based on this limited information, Chevron applied for a large exploration block in mid-1974, and was awarded the acreage in late 1974 ( Figure 1 , Map of the original December 1974 concession, showing areas covering the Muglad and Melut basins, retained at the time drilling began in late 1977. Note the exploratory well locations and boundary of the southern region.).

Figure 1

Because of an almost total lack of geological control, Chevron realized that they had an opportunity to carry out a textbook example of an exploration program (Martini and Payne, 1995). Their first goal was to analyze their block using inexpensive reconnaissance tools and then progressively focus more expensive, higher-resolution tools on the most prospective areas.

The first questions about the area -- the location, extent, and depth of the basin -- were answered with a regional aeromagnetic survey begun in early 1974 and performed in two phases. The first phase consisted of regional coverage of the entire northwest-trending block with flight lines spaced 40 kilometers apart. A preliminary interpretation of this 60,000 line-kilometer survey revealed two separate basins, the Muglad Basin on the northwest and the Melut Basin on the east. A further 21,000 line-kilometers of aeromagnetic data were flown over the deeper Muglad Basin in mid-1974 to improve the resolution of the survey.

The aeromagnetic survey helped to focus the exploration effort, but the focus had to be refined further before expensive seismic operations could be planned. Logistics were an important consideration in this case. The northern half of the Muglad Basin had more favorable access and terrain conditions, and Chevron chose to concentrate in this part of their block (they later relinquished the other acreage). Their next exploration step was a gravity survey, conducted from late 1975 to early 1976. The results of this survey were used to identify major structural trends and anomalies

( Figure 2 , Generalized Bouguer gravity map.

Figure 2

Well symbols depict first six wells drilled in production-sharing agreement area. C.A.R.= Central African Republic.), which further focused Chevron’s efforts. Six hundred miles of reconnaissance seismic were obtained beginning in early 1976, designed to give insight into the nature of the basin, the type of structures present, and reliability of the gravity interpretation (Martini and Payne, 1992). The reconnaissance seismic correlated well with the gravity interpretation and produced regional stratigraphic and structural information.

An initial seismic interpretation indicated the distribution of highly prospective trends. By early 1977, Chevron began shooting more detailed seismic surveys to interpret stratigraphic relationships and structural details for play analysis. Based on seismic data, the predominant plays were assumed to be rotated fault blocks

( Figure 3 , Time-migrated seismic section across Heglig area of southern Muglad block.

Figure 3

Section passes through several productive fault blocks. Times shown are two-way travel times in seconds.), drape folds, and reverse drag folds ( Figure 4 , Time- migrated seismic section across Unity area of southern Muglad block.

Figure 4

Section passes through Unity field. Times shown are two-way travel times in seconds.) developed in an extensional setting. Information on the other play elements was limited because of a lack of outcrops. The rocks were predicted to be non-marine rift-related deposits, with potential reservoir and source elements inferred from seismic facies analysis. Reconstruction of depositional history and depositional systems, however, was difficult without samples (Schull, 1988).

Chevron drilled the first of six wells in late 1977. Although no shows were encountered on the large structure drilled, the well penetrated a thick, organic-rich lacustrine shale. This shale was immature at this location, but the source element now had some encouraging aspects. A second well, Unity #1, tested a major seismic structure, and the presence of oil shows indicated that maturation and migration had occurred in the basin. More importantly, 40 meters of good reservoir rock were associated with the oil shows. This well was abandoned in 1978. The next three wells, drilled in early 1979, were also dry. A sixth well flowed oil from a poor reservoir, but was abandoned as non-commercial in 1979.

Seismic data acquisition continued during the early drilling phase, and well data were being integrated with seismic data as they became available. In-fill seismic data were obtained over the structure drilled at Unity #1, and the revised structure map indicated that the crest of the structure was 13 kilometers south of the #1 location

and up to 100 meters higher. Drilling of the Unity #2 well began in late 1979. When tested in February of 1980, the well flowed oil at the rate of 2,939 barrels per day.

With subsurface information available from seismic data tied to well control, Chevron geoscientists could begin to analyze the structural/stratigraphic relationships in more detail ( Figure 5 , Generalized Muglad Basin structural-stratigraphic cross section.

Figure 5

Well symbols depict first six wells drilled in agreement area.). The well control made depositional systems analysis possible ( Figure 6 , Generalized depositional model depicting non-marine environments operative during the filling of the southern Sudan rift basins.

Figure 6

), and this analysis could be used to predict the distribution of source, reservoir, and seal units throughout the basin. Geochemical samples from the wells made geohistory analysis practical. Seismic acquisition, drilling, and data integration continued as exploration proceeded. Figure 7 summarizes Chevron’s Sudan exploration operations from 1975 to 1985. By 1985, 86 wells were drilled and oil had been discovered in nine areas within the basin.

Figure 7

Lower Indus Basin

A previously explored basin with limited subsurface data can also represent an exploration frontier, if drilling initially indicated poor prospectivity. Exploration ceased in the Lower Indus Basin in the mid-1960s after almost two unsuccessful decades, because existing seismic technology did not allow good data acquisition in a unique geologic environment. When play analysis became possible with new seismic technology in the mid-1970s, exploration resumed and resulted in Union Texas’ discovery of oil in 1981.

The first well in the lower Indus Basin ( Figure 1 , Location map showing the Lower Indus Basin, the Potwar Basin, the Sind, and the Badin Block.) was drilled in 1892, on the surface expression of an anticline (Young, 1995).

Figure 1

This well and a second well drilled on the same structure in 1925 were both dry. A third well was drilled on the anticline in 1956 as part of an extensive basin-wide exploration program carried out by several companies, including Stanvac, Sun, and Hunt. This well tested non-commercial quantities of gas. Five other wells were drilled as part of this program prior to the mid-1960s. Although all these wells were dry, the stratigraphic section encountered in the wells indicated a major Cretaceous depocenter (Young, 1995). The data from this and earlier exploration efforts, including well logs, seismic, and gravity, lay in government files for a decade after the last drilling, until re-examined in 1975 by Union Texas Petroleum. Using the old data, they generated a new play concept.

Existing data indicated both potential reservoir and source units, apparently deposited in a thick deltaic system. These stratigraphic units were equivalent to sequences in Kuwait, which provided a productive analog. Gas shows indicated that maturation and migration had occurred. Trap potential was the questionable element, because existing seismic data failed to penetrate a shallow volcanic unit that masked Cretaceous and older structures. Gravity data provided the only structural control and indicated a belt of northwest-trending gravity highs. These were interpreted to represent Cretaceous structures analogous to adjacent surface anticlines in India. The exploration team developed a geologic model in which northwest-trending structures formed potential traps in a northeast-trending basin with favorable stratigraphic elements. They proposed that the five dry holes were not

sited properly due to the limitations of the seismic data from the 1950s. Their model was accepted by management and Union Texas signed a concession agreement in early 1976.

Union Texas actually had very little useful data to work with. They knew they had a basin with potential favorable reservoir and source units, but they needed more subsurface control to fully develop their exploration concepts. New seismic acquisition was required to penetrate a problematic volcanic unit and for interpretation of the structural and stratigraphic relationships in the basin. A primary goal was to evaluate the proposed gravity high/anticline connection predicted in the preliminary play analysis. A reconnaissance program utilizing improved seismic technology and optimal field parameters showed the true structural relationships within the basin (Young, 1995).

The initial seismic program of 2500 kilometers revealed that, rather than anticlines, the gravity highs were in fact related to thickened sections of high-density carbonate, deposited in structural depressions. At this time the concession agreement required that a well be drilled in order to maintain the acreage, and Union Texas drilled on a poorly-defined structural high in early 1979. This well had an oil show, as did a second well drilled on a better-defined tilted fault block.

Continued seismic acquisition confirmed that the basin structural style was dominated by northwest-trending horst and graben blocks. Encouraged by oil shows and more detailed maps from a tighter seismic grid, Union Texas drilled three more wells. The first two wells had shows, but were non-productive. The third well, Khaskeli #1, the 1981 discovery well for the first oil field in the lower Indus Basin, produced oil from a tilted fault block at approximately 1000 meters. Since 1981, over 30 oil and gas discoveries have been made in the lower Indus Basin.

Marib Basin, Yemen

In July, 1984, Hunt Oil Company tested oil in the Alif #1 well in the Marib Basin, the first oil production in Yemen. This discovery occurred less than four years after the first hint of a potential basin was recognized. In mid-1980, Hunt was alerted to an aeromagnetic survey (flown for mineral exploration) that suggested the presence of an undiscovered sedimentary basin in central Yemen. There were no obvious indications of a deep basin at the surface -- in fact, basement outcrops in the center of the "basin" seemed to indicate shallow basement across the region. The aeromagnetic data, however, suggested 2 to 3 kilometers of sedimentary section, and a salt-cored anticline associated with organic-rich shale was reported within the "basin" area.

Hunt’s explorationists were familiar with Jurassic salt in stratigraphic test wells in an adjacent portion of Saudi Arabia, and recognized a possible correlation with proven Jurassic source rocks elsewhere in the region. The possibilities prompted reconnaissance field trips, which indicated the presence of source rocks with high organic content. Jurassic carbonates exposed along the margins of the basin appeared to have been deposited in a highly unstable environment, suggesting possible rifting during deposition. At this point in the investigation, there were many unanswered questions, but the potential for source and reservoir in a relatively deep basin, together with salt (considered capable of either acting as a seal or of generating structures), surrounded regionally by huge reserves in the Arabian Peninsula, led Hunt to negotiate an exploration contract in July, 1981.

Exploration began in January, 1982, with the acquisition of regional seismic coverage. The first two lines crossed the basin at its widest point, and traversed the northwest-trending axis of the basin. Several more lines crossed the axis of the basin at 20-kilometer intervals. These lines showed both sufficient structuring and a sedimentary section thick enough to encourage continued exploration. The axial line provided the line ties and validated the presence of a long, deep, complex half-graben. Early lines showed a concentration of structures in the southeastern portion of the basin, so the rest of the first seismic program (a total of 1845 kilometers) covered this area with greater density.

Photogeologic mapping and field mapping were conducted during 1982, to further refine surface structural and stratigraphic control. Potential reservoirs were identified as Cretaceous clastics, Jurassic carbonates, Jurassic or Upper Permian sandstones, and Cambro-Ordovician sandstones. Jurassic evaporates were identified as both potential source and seal units. Permian shales were also considered a potential source, although surface exposures were too heavily weathered for effective sampling.

The stratigraphic section identified at the surface provided a provisional stratigraphy for seismic interpretation. The salt and organic-rich shales cropping out within the basin were projected down into the subsurface using the seismic data, where they appeared to blanket a set of tilted fault blocks interpreted to have formed in a transtensional rift environment. The Jurassic carbonates that lie on basement and at the surface were predicted below the salt and shale sections, with possible intervening Cretaceous clastics. Paleozoic clastics underlain by basement were predicted beneath the Jurassic carbonates. Nine separate tilted fault-block structures were identified on the initial seismic program. At this stage, the seismic grid was widely spaced, and further exploration required stratigraphic and structural information from well control to confirm predicted play elements. The Alif structure was chosen for the first exploratory well.

Alif #1 was spudded on January 31, 1984, and provided the necessary information to clarify the producing play. The predicted stratigraphic column was revised, with the addition of approximately 2000 meters of section between the evaporites and Jurassic carbonates seen juxtaposed at the surface on the south flank of the basin. This section, which had not been seen prior to drilling, was comprised of rift deposits that contained source, reservoir, and seal units. The new stratigraphic information was integrated into the seismic interpretation, which indicated that these important play elements extended throughout the eastern part of the basin. The Alif #1 also encountered a thick hydrocarbon column (457 meters) of light oil (40 API) and wet gas, in two zones. The well tested 7800 BOPD on July 4, 1984, and later gas testing recorded 55 million CFGD. Alif #2 confirmed the presence of a thick productive zone, and the exploration program was expanded to define and develop the play.

The exploration program developed along several fronts. Drilling on the Alif structure continued, designed to determine the size of the structure and the potential reservoirs. After drilling 12 wells, Hunt predicted recoverable reserves in the range of 400 to 500 million barrels of oil, with significant qualities of gas. During the Alif evaluation program, a second rig began drilling other large structures. Two of the next three wildcats discovered hydrocarbons. The presence of the reservoir, source, seal, and trap elements was confirmed in these wells. A new seismic acquisition program covered the western part of the basin and filled in the grid in the eastern area, as Alif appraisal continued.

Hunt’s contract required an acreage relinquishment at this stage so the exploration program also had to evaluate the entire basin in enough detail to allow Hunt to let go 25 percent of their concession. To this end, they drilled several stratigraphic test wells around the margins of their concession prior to relinquishing a horseshoe-shaped block of acreage. Two additional plays identified from seismic were also investigated. A carbonate section in a tilted fault block yielded CO2, and an inferred fan complex proved non-commercial.

When the development drilling at Alif confirmed a major oil accumulation, Hunt realized the necessity of building a pipeline to transport the oil from interior Yemen to the Red Sea coast. This pipeline, which runs for over 400 kilometers from the Alif field to the coast, has a capacity of 400,000 BOPD. Hunt chose to sign a joint venture agreement with Exxon during this phase of development.

By 1992, Hunt had shot more than 15,000 kilometers of seismic within their concession and drilled over 170 wells. Nine fields are currently producing, and reserves (primarily gas) have been identified in several other structures. The last "wildcats" were drilled in 1992, and exploration continues with 3-D seismic acquisition and interpretation. Several other companies have become involved in Yemen since Hunt pioneered exploration in the area, and at least two other major accumulations have been found.

Hunt’s early recognition of a viable play, and a well-planned exploration program, can serve as a model for a frontier play analysis and subsequent development. This discovery also has regional implications for oil exploration. The Jurassic rifting that affected Yemen almost certainly affected adjacent areas in eastern Africa. Several companies have attempted to duplicate Hunt’s Yemen success in those areas.

Deep Basin Trap, Western Canada

The January, 1976 discovery of the giant Elmworth gas field in the Deep Basin area of the Alberta Basin, western Canada, demonstrates that creative play analysis can identify significant new reserves in basins considered to be well-explored. By the early 1970s, the Deep Basin area had been abandoned by virtually every large oil company in Canada, because the conventional wisdom at the time considered the area to be unfavorable for hydrocarbon entrapment (Hatley, 1995). In 1973, geologists at the newly-formed Canadian Hunter Exploration began re-examining the Alberta Basin in western Canada using the premise that low-porosity (< 15 percent porosity) Mesozoic gas reservoirs similar to those developed in the United States (e.g., the San Juan Basin, 7-10 percent porosity) might exist in the basin, although they had never been recognized. In fact, in 1974, no gas was being produced in Canada from sands with less than 13 percent porosity.

Canadian Hunter recognized that two new factors encouraged the search for low-porosity gas reservoirs. First, new hydraulic fracturing techniques were being applied to successfully boost production from tight gas sands in the United States. Second, the price of gas had begun to increase, making previously marginal reserves and remote areas more economic. In addition, there was a vast catalog of data from thousands of wells within the Alberta Basin (several billion dollars worth of exploration data) that had never been systematically analyzed with low-porosity Mesozoic gas reservoirs as a target. The objectives for most of these wells had been Paleozoic, and the Mesozoic gas-bearing section was often completely ignored because of the very low price of gas. Only a few small, stratigraphic traps had been identified in the shallow shelf east of the Deep Basin area.

Canadian Hunter analyzed over 5000 electric logs in a regional survey of the Alberta Basin, looking specifically for tested gas shows and for potential productive zones identifiable by porosity and high resistivity (>20 ohms). They constructed several 500-kilometer-long cross sections across the basin showing the entire geologic section ( Figure 1 , Cross section of Alberta showing gas-saturated sands of the Deep Basin.), and plotted maps showing wells with gas indications (Masters, 1995).

Figure 1

Masters’ regional map of gas shows indicated a significant correlation with the San Juan Basin analog. In the productive San Juan Basin, gas occurs in a syncline, and a 500-meter thick, gas-saturated productive zone grades updip into a mixed gas-water transition zone that finally produces only water. The trap ( Figure 2 , Physical forces holding gas in place are not fully understood.

Figure 2

Downdip water flow, "water block" or both appear to be involved.) appears to be formed by a combination of hydrodynamics and a gas permeability barrier created by water saturation in low-permeability rocks (Masters, 1992). This same relationship existed in the Alberta Basin. In the shallow eastern shelf, the section was typically water-bearing, with scattered small, stratigraphic pinch-out gas traps. There was a zone of wells with both gas and water recovery, and finally a zone in the Deep Basin with only gas recoveries and electric log shows (>20 ohms resistivity). Figure 3 is an isopach map of the net thickness of the gas-saturated Mesozoic section in the Alberta Basin, which exceeds 10,000 meters.

Figure 3

The zone is over 640 kilometers long and an average of 96 kilometers wide (Masters, 1979).

Several hundred of the wells analyzed in the study were in the Deep Basin, many of which indicated significant, overlooked gas reserves (Masters, 1995). Canadian Hunter prepared a suite of "Bcf-per-section" maps based on a well-by-well calculation of recoverable reserves determined from well log parameters. A concentration of 85 wells (all previously designated dry holes) in the vicinity of the village of Elmworth showed BCF-per-section values ranging from 8 Bcf to over 25 Bcf, covering an area 160 kilometers long and 80 kilometers wide ( Figure 4 ) This was the target of Canadian Hunter’s first drilling program, and their second well was the discovery well for the giant Elmworth Field.

Figure 4

Canadian Hunter modeled a huge potential play based largely on analogy. They felt confident that new technology would allow them to produce the play, especially with improved prices. They applied well-log analysis on a vast regional scale, which was not typical of the rest of the industry, and identified a prolific new play that hundreds of wells had bypassed. Eighteen years later, Elmworth Field has produced nearly 2 Tcf of gas, and has probable reserves of 5.6 Tcf. The tight gas play has also been extended into British Columbia, and has the potential to be developed in the more structured foothills of the Canadian Rockies (Masters, 1995).

Eolian Stratigraphic Traps, San Juan Basin, USA

In the early 1970s, the San Juan Basin was already recognized as a prolific oil and gas province, with production from several Cretaceous sandstones in both structural and stratigraphic traps. Only one field, the Media Field, had been discovered in the Jurassic Entrada Sandstone in 1953, where development drilling between 1969 and 1974 resulted in production exceeding 700,000 bbls. Wells in this field, however, produced high water volumes making the play marginally economic. The development of high-volume, down-hole pumps capable of lifting large quantities of fluid to the surface, and an increase in the price of oil justified the exploration for similar fields in the basin (Vincelette and Chittum, 1981).

The first step in the exploration program was to characterize the Media Field. The trap was originally presumed to be structural, having been drilled on a seismic nose. Renewed analysis revealed that the closure appeared to rely on topographic relief on top of the Entrada Sandstone. The Entrada thickened 30 meters over the crest of the field, and the overlying Todilto Formation thinned correspondingly ( Figure 1 , Stratigraphic cross section, Media Field.

Figure 1

Cross section was constructed using Cretaceous Sanostee marker as horizontal datum, and drawing base of Entrada Sandstone parallel with Sanostee. For easier viewing, cross section is shortened vertically approximately 1,600 ft (488 m) from Sanostee to base of Morrison Formation.). Previous workers had modeled the Entrada as the remains of an extensive dune field that became the site of a large saline lake in which the limestone/anhydritic Todilto was deposited. A new interpretation suggested that petroleum accumulation at the Media Field was controlled by the topographic relief of a buried sand dune, overlain by lacustrine source rock. This new model was the basis for renewed exploration for Entrada stratigraphic traps in the San Juan Basin.

The next step in play analysis, mapping the basin-wide distribution of the play elements, required a regional well-log analysis of a variety of factors, including the presence of anomalous Todilto-Entrada thicknesses, oil shows, porosity distribution in

the Entrada, and source-rock potential of the overlying Todilto Formation (Vincelette and Chittum, 1981). Figure 2 (Map shows wells which penetrated the Entrada Sandstone in the San Juan Basin.

Figure 2

All data are based on information available prior to initiation of exploration program in 1974.) summarizes the results of the study of porosity and thickness variations. The Entrada is tight in the northwest portion of the San Juan Basin due to compaction and silica cement. Wells in the southeastern part of the basin show anomalously thick Entrada and/or anomalously thin Todilto ( Figure 3 , Regional stratigraphic cross section illustrating thickness changes in the Entrada Sandstone and the overlying Todilto Formation.

Figure 3

Base of Entrada Sandstone is used as horizontal stratigraphic datum.). In addition, 24 percent of the wells in the southeastern area have oil shows in the Entrada, in contrast to few shows elsewhere in the basin. Figure 4 (Map showing results of Todilto Formation source-rock analysis and pyrolysis yields in the San Juan Basin.) summarizes a source-rock analysis of Todilto cuttings and outcrop samples.

Figure 4

The Todilto anhydrite facies has the best source potential, and its zone of maximum oil generation occurs in the southeastern part of the basin. These play summary maps were used to define the area of maximum potential of the play. This area was then chosen for further exploration.

More detailed play analysis demanded more subsurface control on Entrada thickness variations. An experimental seismic program, designed to define a recognizable signature characteristic of the preserved Entrada dunes, was shot over the Media Field and other anomalies. The seismic signal responded both to the thickening in the Entrada and to the thinning of the Todilto with a recognizable signature, thus proving that the topographic relief on the Entrada could be mapped regionally. A reconnaissance seismic program further refined the play maps and identified numerous leads ( Figure 5 , Map showing general location and configuration of Entrada seismic anomalies along southwest flank of San Juan Basin.

Figure 5

Areas shown as dotted pattern indicate Entrada sand thicks or topographic highs. Several anomalies had more than one test, but only the initial wildcat location is shown. Drilling acitivity is as of January 1980.), which were upgraded to prospects by detailed seismic coverage. Vincelette and Chittum (1981) review the expanded exploration program, which has resulted in the discovery of six new Entrada oil fields ( Figure 5 ).

Play Control

Introduction

In order to develop an effective play analysis, we must be able to predict the presence of play elements within a basin. We must also develop an understanding of the spatial and temporal relationships between these elements. The character of each play element is governed by a variety of factors, called play controls. We can build an understanding of the controls on play element development by using models of ancient and modern sedimentary basins. Basin classification schemes provide us with broad-scale models for describing sedimentary basins. The concept of structural styles and tectonic setting provide another perspective from which we can analyze

the relationships between play elements. Depositional systems models provide a third perspective.

There are several relatively simple basin classification schemes in use within the petroleum exploration community. Each of these classifications provides us with models of basin evolution that allow us to predict the development and distribution of the various play elements within a given basin. Because each sedimentary basin is unique, even if formed under very predictable conditions, the best basin classification scheme is not an infallible predictor of trap type, reservoir distribution, source maturation, etc. For this reason, we recommend using a relatively simple basin classification scheme, one that gives us some basic guidelines for what to expect but leaves us with maximum flexibility to interpret and integrate data into our models.

Basin Classification

We classify a basin in order to place it in context with similar sedimentary basins. We can then use these similar basins, whether productive or non-productive, as predictive analogs to help us analyze our basin. Most of these classification schemes separate basins according to the type of crust on which they form -- continental, intermediate, or oceanic -- and to the type of tectonic regime under which they develop -- compress ional, wrench, or extensional. Some basin classification schemes emphasize depositional events during basin formation, while others emphasize tectonic events. Some emphasize factors controlling hydrocarbon accumulation, while others completely ignore the issue. In this text, we will highlight those schemes that address elements involved in play analysis.

Selley and Morrill (1983) present a particularly useful basin classification scheme, shown in Table 1. Let’s consider several examples of the types of basins defined in this classification and examine important aspects of each basin type, including the play elements we have previously mentioned, as well as depositional history, geothermal gradient, risks, and typical reserves.

Crust Type Tectonic Setting Basin Type

Continental Crust Cratonic: Extension & Compression

Interior & Foreland

T

E

R

T

I A

R

Y

Continental Crust & Intermediate Crust

Divergent Margins

Rift & Pull-ApartOpen Down warp

Intermediate Crust Convergent Margins

Closed, Trough

Fore-arc Back-arc Non-arc Collision

D

E

L T

A

Table 1: Classification of sedimentary basins.

Rift basins typically form on continental crust under conditions of tension and divergence. Productive rift basins include the Rhine Graben in Europe, the Sirte Basin in Libya, the Red Sea Rift in the Middle East, and the Viking Graben in the North Sea. The Viking Graben ( Figure 1 ) which shows the half-graben geometry typical of many rifts, produces mainly from tilted fault blocks in pre-rift Jurassic sandstones.

Figure 1

Upper Jurassic post-rift shales provide source rocks that also act as seals. Hydrocarbon production also comes from several post-rift carbonates and deep-water sandstones. The Viking Graben, with its three separate plays, is a good productive analog for a predominantly clastic rift basin. Table 2 summarizes the major characteristics of rift basins.

Rift Basin

Distinguishing Features:

down dropped graben over continental crust; dormant divergence

Depositional History:

prerift rocks sedimentary, metamorphic or granitic; post-rift fill is restricted facies, initially non-marine that may become marine (either clastic or carbonate-prone)

Reservoir: equally sandstone or carbonate; of pre- and post rift cycles

Source: Overlying or lateral facies shale

Cap: basin wide evaporites or thick shales

Trap: horst block anticlines; combination traps related to high blocks; tilted fault blocks

Geothermal Gradient: normal to high

Hydrocarbons: highly facies-dependent (paraffinic with sandstones; aromatic with carbonates); low to average gas

Risks: small trap size; too high gradient; source shale development

Typical Reserves: <0.5-3.0 billion bbl hydrocarbon/basin

Table 2: Characteristics of rift basins.

Rifts may become pull-apart basins if opening continues and an ocean basin forms. These basins are asymmetric, and have a continental sediment source. Examples of productive pull-apart basins include the Senegal and Gabon basins offshore west Africa, the Northwest Shelf Basin offshore Australia, and the Hibernia Basin offshore Newfoundland. Figure 2 ,

Figure 2

a generalized cross-section through the Gabon Basin, shows the basin contains potential hydrocarbon accumulations throughout its sedimentary section, including rift-stage fault traps, drift-stage shallow-marine sediments draped over high blocks, and salt tectonic-related traps in post-separation stage clastics. The Gabon Basin is typical of many pull-apart basins, and serves as a good analog. Table 3 summarizes the characteristics of pull-apart basins.

Pull-Apart Basin (passive margin, divergent margin)

Distinguishing Features: Coastal half-grabens down-faulted seaward; intermediate crust result of ocean-floor spreading

Depositional History: non-marine rift stage sediments; restricted facies (carbonates, evaporites, black shale) in early separation; prograding clastic wedge in late separation stage

Reservoir: sandstone in all 3 stages; some limestone in early separation stage

Source: overlying and interfingering shale

Cap: shale or evaporite

Trap: horst block, salt flow, roll-over and drape anticline; stratigraphic and combination

Geothermal Gradient: below average in marine stages

Hydrocarbons: rift stage has paraffinic, intermediate gravity crude; more aromatic, higher gravity in separation stage; gas prone

Risks: kerogen maturation; biodegradation; pre-separation source shales; post-separation reservoirs

Typical Reserves: 2-3 billion bbl hydrocarbons/basin (none fully developed)

Table 3: Characteristics of pull-apart basins.

Selley and Morrill (1983) propose a special class for sedimentary basins which are down warps into small oceans because their sediments and petroleum characteristics are often very different from other basins types to which they are genetically related. The down warp basin classification includes basins which are related to the opening of small ocean basins that did not continue to spread, such as the Gulf Coast Basin offshore southern North America. Figure 3 , a generalized cross section through the Gulf Coast Basin, shows Triassic rift-related sediments overlain by evaporites and

several kilometers of shallow-marine clastics and carbonates.

Figure 3

Three plays predominate within the basin: salt domes activated by rapid Tertiary subsidence and infilling, normal fault traps in Tertiary clastics, and stratigraphic pinch-outs along the basin margins.

Small, closing-ocean basins also fall into the down warp classification. The Arabian-Iranian Basin ( Figure 4 ) is a good example, with the Arabian Gulf being a remnant of the Tethyan Sea.

Figure 4

Convergence of the Arabian and Asian plates produced restricted basin conditions favorable for petroleum formation, as well as numerous types of traps. Production from carbonates predominates, but clastic reservoirs are productive as well. Shales and shaly limestones deposited during transgressions are the major source rocks, and widespread evaporites form excellent seals. Thrust-related anticlinal traps define the play on the eastern margin of the basin, while subtle folds related to drape and salt diapirism define the play on the western margin. Table 4 summarizes the important characteristics of down warp basins.

Rift Basin

Distinguishing Features:

down dropped graben over continental crust; dormant divergence





Cap: basin wide evaporites or thick shales







The down warp basin classification appears to group together basins formed under dissimilar conditions, but the conditions associated with the filling of small ocean basins are similar enough to justify this approach (Selley and Morrill, 1983). Some down warp basins may appear in other classifications. For example, the Gulf Coast may be classified as a pull-apart basin, and the Arabian-Iranian Basin may be called a convergent-margin basin. The choice of classification may depend on the amount and type of data available, and may change as more data become available.

Selley and Morrill also group together the several kinds of convergent margin basins: the fore-arc, back-arc, non-arc and collision basins. Back-arc and fore-arc basins form near subduction zones that have developed island arcs. Fore-arc basins may be poor petroleum prospects because of low heat flow and poor-quality reservoir facies, while back-arc basins may be attractive petroleum prospects since they often have high heat flow and favorable clastic reservoir facies. Figure 5 illustrates the convergent margin basins developed in the Sumatra and Mentawai basins of Indonesia.

Figure 5

Non-arc basins form where converging plates move past each other obliquely, forming wrench fault zones that deform through a combination of both transcurrent fault movements and local block faulting. Figure 6 shows a cross section through the Los Angeles Basin, a non-arc basin related to oblique convergence.

Figure 6

This basin is filled predominantly with clastic turbidity current deposits. Compressional anticlinal traps form the major play. Source rocks are the Miocene deep-water shales which have matured because of high heat flow within the basin. Table 4 summarizes the characteristics of convergent-margin basins.

Basin-forming tectonics are a key element in the basin classification scheme presented by Kingston et al. (1983). This classification scheme is valuable as a guide to basin evolution. The terminology differs from that of Selley and Morrill (1983), as does the classification of a few types of basins, but they are generally compatible.

This scheme presents models for the development of several important basin types. For example, an interior fracture basin (rift basin) forms on continental crust, either in the interior of present plates or at the crustal margins of old continental plates. Figure 7 illustrates the three main stages in the development of an interior fracture basin.

Figure 7

In Stage 1, block faulting produces an interior graben system and depressions are filled with non-marine clastics. In Stage 2, the graben system deepens to form a basin. Marine waters invade and marine beds are deposited. Sandstone is deposited and limestone reefs form over highs, while shales accumulate in lows. Stage 2 ends with salt deposition. In Stage 3, non-marine deposition occurs as the basin fills and deposition overcomes subsidence. Faulting is not generally observed; sag and subsidence predominate.

A margin sag basin (pull-apart basin) forms at the edge of continental crustal blocks, under extensional conditions. Figure 8 illustrates the four main types of margin sag basins, classified according to the type of sedimentary basin fill.

Figure 8

Margin sag basins that later become involved in continental collision form a fold belt on the former "seaward" side of the basin, and become a type of foreland basin.

Kingston et al. take their classification scheme one step further than most by introducing the concept of tectonic modifiers, which are fold belts or reactivated wrench faults that affect basins after their formation. Understanding how basins can be modified can help us to unravel complicated basins, identify their original basin type, and fit them into appropriate interpretation models. Figure 9 illustrates examples of simple divergent margin basins affected by folding and reactivated wrench faulting.

Figure 9

Figure 10 illustrates the idea of polyhistory or successor basins, in which several basin types succeed each other, and are affected by basin-modifying tectonics during their evolution.

Figure 10

This example illustrates some of the complications inherent in basin classification. This basin classification scheme is detailed and somewhat unwieldy, but the concepts and models presented are valid and useful.

We recommend familiarity with several different classifications schemes. Each scheme addresses important play elements and provides valuable models and examples, which we can use when making decisions regarding basin classification.

Structural Styles

Once we have classified our sedimentary basin, we can then expect a dominant structural style or limited set of structural styles to exist within that basin, which we can then use to predict play types in the basin. The concept of structural styles is a powerful predictive tool, one which we can use to great advantage in petroleum exploration, especially in frontier areas. The connection between structural style and structural trap formation is obvious, but stratigraphic and combination traps also fit into the structural style framework in predictable ways. Maturation and migration processes differ from one structural style to the next, as do depositional processes.

Structural styles interpretation, introduced by Dahlstrom (1970), suggests that only a limited suite of structures will exist in a given geologic environment. Harding and Lowell (1979) emphasized the application of structural styles concepts to petroleum exploration, and these applications were developed further by Bally and Oldow (1983) and Lowell (1985).

We classify structural styles first based on the tectonic regime in which they form, whether Compressional, strike-slip (wrench), vertical, or extensional. We further subdivide each of these classifications based on whether basement is involved (basement-involved structures) in the deformation or whether only the sedimentary cover is involved (detached structures). Table 1 lists the structural styles and their various plate tectonic habitats, as proposed by Harding and Lowell (1979). Notice that many of the plate tectonic habitats sound suspiciously like types of basins mentioned in our discussion of basin classification above. Remember that our basin classification scheme separates basins according to tectonic setting.

Dominant Typical Plate Tectonic Habitats

Structural Style

Deformational Force

Transport Mode Primary Secondary

ExtensionalExtensional fault blocks

Extension High to low-angle divergent dip slip of blocks and slabs

Divergent boundaries: 1. Completed rifts 2. Aborted rifts aulacogens

Convergent boundaries: 1. Trench outer slope 2. Arc massif 3. Stable flank of forehand & fore-arc basins

4. Back-arc marginal seas (with spreading) Transform boundaries: 1. With component of divergence 2. Stable flank of wrench basins


Extension Subhorizontal to high-angle divergent dip-slip of sedimentary cover in sheets, wedges and lobes

Passive boundaries (deltas)

Salt structures Density contrast Differential loading

Vertical and horizontal flow of mobile evaporites with arching and/or piercement of sedimentary cover

Divergent boundaries: 1. Completed rifts and their passive margin sags 2. Aborted rifts; aulacogens

Regions of intense deformation containing mobile evaporite sequence

Shale structures Density contrast Differential loading

Dominantly vertical flow of mobile shales with arching and/or piercement of sedimentary cover


Regions of intense deformation containing mobile shale sequence

Gravity structures Slope instability Differential loading

Down slope gliding on decollement


Convergent boundaries: 1. Trench outer slope 2. Fore-arc basins 3. Back-arc basins

CompressionalCompressive fault blocks & basement thrusts

Compression High to low-angle convergent dip slip of blocks slabs, and sheets

Convergent boundaries: 1. Foreland basins 2. Organic belt cores

Transform boundaries: (with component of convergence)

3. Trench inner slopes and outer highs

Decollement thrust-fold assemblages

Compression Subhorizontal to high-angle convergent dip slip of sedimentary cover in sheets and slabs

Convergent boundaries: 1. Mobile flank (orogenic belt) of forelands 2. Trench inner slopes and outer highs

Transform boundaries: (with component of convergence)

Strike Slip Basement InvolvedWrench fault Shear couple Strike slip of sub

regional to regional plates

Transform boundaries:

Convergent boundaries: 1. Foreland basins 2. Orogenic belts

3. Arc massif Divergent boundaries: 1. Offset spreading centers

VerticalBasement warps: arches, domes, sags

Multiple deep-seated processes (thermal events, flowage, isostasy, etc.)

Sub vertical uplift and subsidence of solitary undulations

Plate interiors Divergent, convergent, and transform boundaries Passive boundaries

Table 1: Structural styles and their various plate tectonic habitats.

Referring to Table 1, we see that convergent boundaries are the primary tectonic habitat of Compressional structural styles, both basement-involved and detached. Basement-involved reverse faults occur either in foreland basins, which form between the volcanic arc and craton, or in fore-arc basins during continental collision. Although these faults also occur in association with trenches, those associated with foreland basins contain the greatest oil reserves.

The geometry of basement-involved reverse structures ranges from high-angle reverse faults to low-angle basement thrusts ( Figure 1 , Four proposed styles of basement faulting in the Rocky Mountain foreland region.).

Figure 1

Seismic evidence supports this diversity of geometry, which develops because the sedimentary cover deforms differently than the basement ( Figure 2 , Two interpreted seismic profiles from Wyoming showing expression of rigid basement uplifts.

Figure 2

[a] Southwest Wind River fault. [b] Casper Arch thrust.). A variety of hydrocarbon traps can be associated with basement uplifts ( Figure 3 , General hydrocarbon trapping possibilities associated with basement block uplifts.) as well as with low-angle basement thrusts.

Figure 3

Many foreland basins are former pull-apart basins subjected to convergent forces, like the Andean foreland basins. Convergence can create new trap structures, but maturation and timing must be favorable for hydrocarbon accumulation to occur.

Detached Compressional structures are common in fold-thrust belts at the edges of foreland basins, and within subduction complexes in fore-arc basins. The former tectonic habitat may be very important for petroleum accumulation, with a combination of favorable play elements ( Figure 4 , Detached compressional structures [over thrusts].). The latter habitat is generally unfavorable, because of low heat flow and poor reservoir conditions.

Figure 4

Strike-slip styles occur in a variety of tectonic habitats along plate boundaries

( Table 2 ). Since plate motion is usually oblique to plate boundaries, components of divergence and convergence may alternate along plate margins. These tectonic regimes create the complex trap geometries shown in Figure 5 (Trapping possibilities associated with wrench faults.).

Figure 5

Rapidly forming compressional uplifts can occur near rapidly subsiding extensional basins, yielding very high sedimentation rates and good clastic sources. The Los Angeles and Ventura basins of California are examples of small, prolific oil provinces on strike-slip margins.

Setting System

Non-marine Alluvial fan Braided stream Meandering stream Eolian Lacustrine

Marginal Marine Deltaic Beach Estuarine Barrier Island Tidal flat Lagoon

Shelf Sand sheet Sand ridge

Slope Canyon fill

Slope Fringe Submarine fan Slope-apron deposit

Basin Plain Basin shale

Table 2: Clastic depositional settings and systems.

Trap geometries can change during the evolution of wrench basins ( Figure 6 , Evolution of wrench or shear basins.

Figure 6

LL=lateral. Development of convergent plate wrench in three stages: LL-3, 2, 1. L3FB shows final step in many wrench basins-that of a foldbelt caused by wrenching. This is the final step in a process that creates a basin and finally destroys it. After basin initiation [LL-1}, movement may cease in any succeeding stage. The basin may also stop its strike-slip mode and become a polyhistory basin.). In the initial stages of basin development, hydrocarbons may be trapped in block-faulted structures, while in later stages of basin development, the predominant trap may be in en echelon folds away from the main wrench zone. Continued deformation may break up large structures, destroying or reducing the size of related oil fields. Associated local-scale strike-slip faulting, such as the tear faults commonly identified in fold-thrust belts, can also be important in trap formation. In this tectonic habitat, the structures are

usually regional in extent and exert important controls on sedimentation. So, sedimentation type will also change as the basin develops.

Vertical structural styles include basement arches, domes, and sags, are usually regional in extent, and occur primarily in plate interiors ( Table 2 ). Reactivation of basement structures during basin formation often controls trap formation within these basins, both by creating faults and by controlling sedimentation in overlying sections. Because of the size and longevity of such basins, favorable depositional environments, such as reef trends and stratigraphic pinch-outs, may extend over large areas. Figure 7 illustrates a variety of potential trap configurations associated with vertical structural styles.

Figure 7

From Table 2, we can see that rifts and passive margins are the primary tectonic habitat of extensional structural styles, both basement-involved and detached. In extensional regimes, growth faults and salt structures are important to petroleum exploration since they can affect both trap formation and deposition. Growth faults are most common along passive boundaries where clastic sedimentation rates are high and sediments remain unconsolidated to depths of several kilometers. The characteristics of growth faults include syndepositional offset and a listric, concave-upward shape. Extensional faults are frequently associated with secondary synthetic

and antithetic faults. Figure 8 shows common geometries for growth faults, and the variety of associated trap styles.

Figure 8

Salt structures are found in many tectonic settings, including passive margins, that are successors to the marginal rift and small, restricted-ocean basins where evaporites form. Salt structures appear to form early with non-uniform sediment loading, and later may form diapirs or salt domes when salt is buried deeply enough to flow plastically. Salt diapirs create a variety of fault, fold, and synsedimentary stratigraphic traps ( Figure 9 ,

Figure 9

Common types of traps associated with salt structures: 1) anticline; 2) graben caused by extension; 3) porous cap rock; 4) flank sand pinchout; 5) overhang; 6) non-overhanging wall of structure; 7) angular unconformity; 8 and 9) normal faulting along flank of structure.). Salt mobilization may also be related to the reactivation of basement faults, or may occur along convergent boundaries where intense deformation mobilizes evaporite sequences.

Extensional block faulting offers many trapping opportunities ( Figure 10 , Major trapping possibilities associated with extensional block faulting.

Figure 10

BC=basement complex.), and its occurrence within some larger rift basins, such as the North Sea Basin, places these traps in close association with favorable source beds, migration conditions, and seals. High heat flow is also typical in the early stages of rifting, creating ideal hydrocarbon-generating conditions.

Rift and passive margin tectonic habitats illustrate the often close relationship between tectonic habitat, structural style, and sedimentation. Figure 11 shows the evolution of an idealized pull-apart basin along the Atlantic continental margin, from the initial rift phase during the Triassic to the present.

Figure 11

A typical stratigraphic section would begin with non-marine clastics, such as alluvial fan, braided stream, and possibly lacustrine deposits. Mafic volcanics are also commonly interbedded with the clastics. This sedimentary section, which fills block-faultÐcontrolled basins within the rift, is overlain by evaporites deposited as the ocean invades the rift during plate separation. The evaporites may occur in separate block-faulted basins or may form broad regional deposits as the underlying rift architecture is buried. The passive margin environment is established as plates continue to separate, and thick wedges of clastic sediment begin to build oceanward and bury the block-faultÐcontrolled terrain. Detached extensional structures become dominant, although basement control may be significant as continued subsidence and loading reactivate old rift structures. A variety of depositional systems can develop along passive margins. Several potential play types occur within both the rift and passive margin sections of the basin ( Table 3 ).

Rift Basin

Distinguishing Features: downdropped graben over continental crust; dormant divergence





Cap: basinwide evaporites or thick shales







Depositional Systems

In addition to considering basin classification and structural style, we can predict many of the play elements within the framework of a given depositional system. A depositional system consists of an interconnected set of depositional environments that exist at a given stage of basin filling. Several different depositional systems may follow one another as a basin evolves, and certain systems are predictable within the basin classification and structural styles frameworks mentioned above.

Basin-filling sedimentation affects every play element in our analysis. Our models must show the relationships between sediments within a basin, so that we can predict the patterns within our basin. We broadly classify sediment accumulations as clastic, carbonate, or evaporite. Within this broad framework, we can define a group of sedimentary settings based on both modern and ancient accumulations. We further subdivide these settings into depositional systems, and finally into sedimentary facies.

There is some confusion in the literature about terminology. For example, a delta may be classified as a depositional system, a depositional setting, or a depositional environment. We have chosen the hierarchy shown in Table 1, Table 2, and Table 3 as a compromise. In this hierarchy, a depositional setting is a broad geographic location within a basin, e.g., a non-marine area versus a marine area. A depositional system is the stratigraphic equivalent of a modern depositional environment, which is distinct chemically, physically, and biologically from surrounding environments. A depositional system consists of an assemblage of interconnected facies formed within subenvironments. For example, within the marginal marine setting, a deltaic depositional system can be subdivided into delta plain, delta front, and pro-delta facies.

Setting System


Marginal Marine Deltaic Beach Estuarine Barrier Island Tidal flat Lagoon


Slope Canyon fill




Setting System

Non-marine Lacustrine Dune Caliche Cave deposit

Marginal Marine Beach Tidal flats (sabkhas) Organic swamp

Shelf Mud banks Tidal delta Muddy shelf sand Patch reef Sand sheet Shelf basin

Shelf Margin Ecologic reef Oolitic/skeletal sand body

Patch reef Beach

Foreslope Turbidite Pinnacle reef Skeletal debris fans

Basin Basin mud Basin chalk Turbidite sand

Table 2: Carbonate depositional settings and systems.

Setting System

Non-marine Lacustrine sabkhas Interdune sabkhas

Marginal Marine Sabkha Salina

Marine Platform Basin-wide

Table 3: Evaporite depositional settings and systems.

In order to perform an effective play analysis, we must understand the potential depositional settings and systems in our area of interest so that we can analyze the accumulation and geometry of reservoir, source, and seal units. We need to understand the relationships among these elements because they affect stratigraphic and structural traps, as well as maturation and migration. The best approach to analyzing depositional systems within a basin is to construct paleogeographic maps to show the distribution of systems and facies during deposition. We will present several models that show the relationships among depositional systems in non-marine, marginal marine, and marine settings. Although familiarity with such models will help in constructing paleogeographic maps, it is important to remember that tectonics, sediment supply, climate, and other variables differ in each basin.

Clastic Depositional Systems

Clastic sediments can accumulate in all depositional settings within a basin

( Table 1 ). In the non-marine setting, five major systems exist: alluvial fan, braided stream, meandering stream, eolian and lacustrine ( Figure 1 , The five general non-marine environments relevant to petroleum exploration.). The alluvial fan, braided stream, and meandering stream systems grade one into the other as we move progressively downstream, from eroding uplands to a marginal marine setting.

Figure 1

Of these three systems, the alluvial fan system, where the elements of reservoir, source, and seal are high risk, is the least prospective petroleum target. Reservoir units are laterally discontinuous and poorly sorted. Shales are equally discontinuous and thin, precluding extensive source and seal accumulation. Alluvial fans are, however, a good indicator of proximity to the basin margin and are sensitive records of the rate of tectonic uplift. They grade basinward into more favorable braided stream deposits, which provide better reservoir rock characteristics, including improved sorting and continuity. Braided stream deposits suffer from the same constraints on source and seal deposition that affect the alluvial fan system.

Setting System


Marginal Marine Deltaic Beach

Estuarine Barrier Island Tidal flat Lagoon


Slope Canyon fill




The meandering stream system occurs closest to the marginal marine environment. While reservoir sorting may improve, reservoir continuity is poorer here than in the braided stream system. The geometry and distribution of individual channel sands are also difficult to predict. This system commonly grades basinward into the deltaic systems of the marginal marine environment, and continuity between porous and permeable sediments deposited in the two systems may provide migration pathways for oils sourced in marginal marine and marine environments. Overlying source rocks and seals may also be deposited during marine transgression.

Eolian deposits may accumulate in a variety of locations, adjacent to any of the deposits in the other non-marine systems. They often form sheet-like units of clean, well-sorted quartz sandstones that are excellent reservoirs and may be good pathways for hydrocarbon migration. Once again, our primary concern is the lack of associated source and sealing units, except where dunes form near the marginal marine or lacustrine settings and interfinger with more shale-prone systems. Tectonics may also play an important part in juxtaposing source and seal units with eolian sands, as in the Rotliegendes play in the southern North Sea Basin. Here, Lower Permian non-marine sediments, including eolian sands, overlie truncated Pennsylvanian coal measures eroded prior to subsidence of the North Sea Basin. Marine infilling led to the deposition of overlying evaporitic seal units. Trap formation occurred next, followed by maturation and gas migration as extensional tectonism and subsidence continued.

Lacustrine systems also occur in a variety of locations in the non-marine setting. Only large lakes, however, have the potential to generate significant quantities of hydrocarbons, and these are typically associated with interior sag and rift basins. Lacustrine systems can produce rich oil-prone source rocks that may interfinger with the reservoir units of other non-marine systems, such as dunes, braided streams, and alluvial fans. A large lake may also form a self-contained, hydrocarbon-generating and trapping system (Montgomery and Selley, 1984). Figure 2 illustrates the major facies in the lacustrine system of Lake Uinta, which comprise potential reservoir, source, and sealing units.

Figure 2

Sandy shoreline deposits pinch out into basinal source shales to form stratigraphic traps with reserves that exceed 100 million barrels. Lacustrine shales are also very important source units in a series of large rift basins of Cretaceous to early Tertiary age in eastern China.

So, while non-marine clastic depositional systems can host major hydrocarbon accumulations in many sedimentary basins, their remoteness from favorable source and sealing unit depositional environments makes them high risk. Tectonic events can juxtapose non-marine reservoirs with sources and seals, or these systems may interfinger with favorable facies in marginal marine systems. Only the lacustrine system is favorable for deposition of a complete suite of source, reservoir, and seal units, and only in large lakes. Some non-marine systems are good tectonic regime indicators. Thick alluvial fan and braided stream deposits are often related to uplift along basin margins, combined with rapid basin subsidence common in active rift environments. Large lacustrine systems with thick basinal shale deposits indicate high subsidence rates relative to sedimentation rates. Extensive eolian deposits are often indicative of an arid to semi-arid climate, as are certain types of lake deposits. While non-marine sediments may not be the most prospective, they are important indicators of climate, tectonics, and basin geometry.

Depositional systems in the marginal marine setting ( Table 1 ) provide more favorable sites for hydrocarbon accumulation, because potential reservoir facies are

commonly adjacent to potential source and sealing facies. Figure 3 shows the relationships between several interconnected depositional systems in the Eocene Wilcox Group of Texas.

Figure 3

An idealized model of sandy reservoir facies in the Wilcox Group ( Figure 4 ) shows how barrier island, deltaic (delta front and distributary channel), and non-marine meandering stream system sands interfinger with potential source and sealing facies in the deltaic (prodelta), tidal flat, and marine shelf systems.

Figure 4

We are particularly interested in the deltaic depositional system in the marginal marine setting because it is the thickest and most widespread system (note scale, Figure 3 ). Deltas build outward into a basin by progradation. As one portion of the offshore fills with sediment, deposition shifts to a new location, or lobe. As abandoned lobes subside due to compaction or rising sea level, other systems deposit sediment and build up intricate stratigraphic columns ( Figure 5 , Composite stratigraphic column of recent deposits from Senegal River delta, a wave/current-dominated delta.).

Figure 5

These vertical relationships are good clues to deltaic deposition. Figure 6 (Idealized block diagram and cross sections showing principal environments and facies of a regressive barrier island system.)

Figure 6

illustrates the stratigraphic relationships between the meandering stream system and the deltaic system in the Wilcox Group and the modern Mississippi Deltas, both in cross section and in map view.

Deltaic facies (marsh and swamp) may source updip fluvial systems, and deltaic sediments are redistributed into other systems in the marginal marine setting. Sediment in the barrier island system, the other marginal marine system of most interest to us, is usually supplied by a deltaic system and redistributed by currents. Figure 7 is a model of a barrier island system (regressive), showing the relationships between potential reservoir facies (shoreface, beach-dune, washover fan, and tidal delta sands) and potential source and seal facies (lagoon, marsh, and shelf shales).

Figure 7

Repeated episodes of transgression and regression along a coastline will produce stair-stepped groups of individual barrier island systems ( Figure 8 ).

Figure 8

Depositional systems in the marginal marine setting are influenced by both non-marine and marine controls. The former include river discharge, sediment supply, and provenance, while the latter include waves, tides, and currents. Preserved systems give us clues to climate, and therefore to organic productivity. They also indicate tectonic controls. Steeper slopes associated with active margins tend to restrict coastal deposition and enhance the transport of material into deeper water, while gentler slopes associated with passive (pull-apart) margins allow marginal marine environments to become laterally extensive and continuous (Morrill, 1987).

Within the marine setting, sediments accumulate in the shelf, slope, slope fringe, and basin plain depositional systems ( Figure 9 , Paleogeography during deposition of the Shannon Sandstone, Powder River Basin, Wyoming.).

Figure 9

Tidal currents and storm currents disperse clastic sediments across the shelves, where the sediment accumulates as sand sheets or ridges. Sand moves in large waves within sand sheets, which may cover tens of thousands of square kilometers and be tens of meters thick. Sand ridges can occur in parallel ridges, or as singular ridges related to topographic highs. They may occur on top of extensive sand sheets. Ridges may be 50 km long and up to 3 km wide, with a thickness of several tens of meters. Sheets and ridges are usually encased in shelf shales, which can provide both seal and source. They may also interfinger with marginal marine deposits at the top of the shelf. The Upper Cretaceous Shannon Sandstone of the Western Interior Basin of North America is a classic example of a productive shelf depositional system.

Very little sediment accumulates on the slope, except in channels that carry sediment into the deeper basin. Figure 10 , a model of a canyon-fed submarine fan, shows how the channel fill in the canyon feeds directly into the channeled upper fan facies of a submarine fan.

Figure 10

Slope channel systems may be prospective, e.g. the Rosedale channel sandstone of southern California (Martin, 1963). Slope channel deposits are mainly useful as a guide to the location of the slope-fringing systems deposited along the edge of the slope.

Two sand facies predominate in the slope-fringing system: the submarine fan and the slope-apron facies. The submarine fan facies is sourced by a channel that carries sediment through the slope onto the basin plain, where it accumulates in a fan shape. In the slope-apron facies, sediments lack feeder channels on the slope and also lack channels in the sediments accumulated on the basin plain. Multiple sediment sources feed sand over a non-channeled slope. Figure 11 compares models for a canyon-fed submarine fan and a delta-fed submarine ramp (one type of slope-apron deposit).

Figure 11

These models also show the relationship between marginal marine deltaic sediment sources and the downslope systems. The sands in both submarine fan and slope-apron facies interfinger with basin plain shales that have both source and seal potential.

The slope-fringing system forms in different configurations within different tectonic habitats. Fans formed on pull-apart margins tend to become elongated within the large, relatively flat basin plain, and are usually fed by a single channel. Fine-grained sands and muds predominate because of the distance from sediment source. The modern Mississippi Fan is a good example of an elongated fan. Radial fans typically form along active margins in restricted basins floored by continental crust (Morrill, 1991). These fans, also fed from a single channel, tend to be composed of coarser-grained clastics with a high ratio of sand to shale. Slope-apron deposits also tend to form along active margins and are usually coarser-grained. The areal extent and thickness of slope-fringing systems are highly variable.

Basin-plain systems are dominated by mud deposition, but density currents can distribute sand in the deep-marine environment. The Ramsey Sandstone of the Upper Cretaceous Bell Canyon Formation was deposited on the basin plain by density currents that cut and filled channels in the basin shales. The channels range from 1.5 to 6 km in width and are generally 10 to 25 m thick ( Figure 12 ).

Figure 12

Interbedded organic rich shales within the sands provide source rocks, and basin shales envelope the sand bodies to provide seals.

Carbonate Depositional Systems

Carbonate sediments typically accumulate in tropical and subtropical marine settings where clastic sediment input is small. Carbonates are generated in place by organic activity within a basin, rather than transported into the basin, as are clastics. Waves and currents redistribute carbonate particles through the various depositional systems. Shallow-water carbonate sediments can accumulate to great thicknesses where sediment production keeps pace with subsidence.

Extensive research over the past 40 years, investigating both modern and ancient carbonate deposits, has led to a conceptual model for carbonate deposition. This "carbonate shelf" model is reviewed in detail by Rose (1987). Figure 1 shows the model in profile and map view.

Figure 1

We will discuss only a few of the many depositional settings and systems of the model.

Tidal flat systems in the marginal marine setting are excellent stratigraphic benchmarks, indicating both paleo-sea levels and paleo-shoreline locations (Rose, 1987). They are easily identified due to characteristic sedimentary structures and geometry. They are also important petroleum reservoirs worldwide. Interbedded carbonate and evaporite deposition is common in tidal flat systems, as is synsedimentary dolomitization. Dolomitized carbonates have fine intercrystalline porosity, and associated evaporite layers can serve as seals. We can subdivide an arid tidal flat (sabkha) system into supra-, inter-, and sub-tidal facies; dolomitization is most common in the inter- and sub-tidal facies ( Figure 2 , Idealized tidal flat model combining onlap and offlap features of Andros Island and Persian Gulf tidal flats to show stratigraphic traps and related porous accumulations.).

Figure 2

Trapping conditions occur where intertidal and channel facies pinch out beneath and within relatively impermeable supratidal facies. Additional reservoir development may occur in tidal deltas and channel and barrier island sands that lie seaward of supratidal and continental facies. Giant oil fields in the Permian San Andres Formation of western Texas produce from these two facies, where supra-tidal anhydritic dolomites provide seals ( Figure 3 , North-south cross section showing three reservoirs sealed by impermeable anhydritic facies to the north.).

Figure 3

The shelf margin is another important stratigraphic benchmark in the carbonate shelf model, marking the transition zone from shallow- to deep-water depositional settings. This setting also includes an array of potential carbonate reservoir targets that grade basinward into organic-rich basin plain mud. Two of the most prospective systems, reefs and sand bodies, are the most extensively developed along the shelf margin. Rose (1987) emphasizes the characteristic topographic relief associated with both ecologic and stratigraphic reefs. The former is the result of organisms that produce a rigid, sediment-binding framework that forms a wave-resistant topographic feature. In the latter, topographic relief may be the result of penecontemporaneous inorganic cementation (Rose, 1987). Figure 4 illustrates a modern ecologic reef, showing the environmentally-controlled zonation of the different reef-building organisms across the reef.

Figure 4

Reef-building activity, as well as types of reef-building organism, has varied through time, but we can correlate growth form with environment and can use reef zonation to interpret basinward and landward orientation. Shelf margin sand deposition can be contemporaneous with active reef deposition, or may predominate along the shelf margin when environmental conditions inhibit reef growth.

Shelf margin depositional systems grade basinward into foreslope systems; those with sediments directly derived from reef systems include debris flows, talus, and turbidite deposits. These may form prolific plays, as in the Poza Rica trend, Mexico, where gravity and debris flow sediments of the Tamabra Formation form reservoirs that interfinger basinward with lime mud source and sealing lime. The Golden Lane trend in the Gulf of Mexico lies on the shelf margin setting and is the provenance for carbonate detrital material in the foreslope (Boyd, 1963; Rose, 1987)

Carbonate rocks are far more susceptible to extensive diagenetic changes than are clastic rocks. The subject of carbonate diagenesis is beyond the scope of this module, and a detailed understanding of the carbonate depositional model and carbonate diagenesis is not necessary for most exploration geologists.

Evaporite Depositional Systems

At first glance, evaporite depositional systems ( Table 1 ) may seem to be a non-prospective rock type, but in fact there is a strong association between evaporites and petroleum reserves. Warren (1992) summarizes the reasons for this association:

Setting System

Non-marine Lacustrine sabkhas Interdune sabkhas

Marginal Marine Sabkha Salina

Marine Platform Basin-wide

Table 1: Evaporite depositional settings and systems.

Beds of ancient evaporites overlying or up-dip from porous sediments create impressive seals for any underlying potential reservoirs.

Thick, buried halite units are often remobilized into salt structures which then create potential reservoirs in surrounding sediments.

The subsurface movement of pore waters can leach evaporites in carbonate and siliciclastic matrices, creating areas of secondary porosity.

Evaporite diagenesis releases large quantities of magnesium-rich brine that can alter limestone to sucrosic dolomite -- an excellent potential reservoir.

In addition, Kirkland and Evans (1981) document the close association between source rock deposition and evaporite depositional settings.

The importance of the marginal marine sabkha to petroleum exploration was mentioned in Section 3.3.2. Warren (1992) lists potential sabkha-associated reservoirs and cites several examples of ancient sabkha-related production. He also presents a thorough discussion of the sabkha depositional model and synsedimentary dolomitization.

Evaporites, in addition to acting as potential source and sealing units and potential diagenetic agents, also mobilize within their surrounding sediments and create hydrocarbon traps.

Play Elements

Source

Without an adequate source rock, the other play elements become irrelevant.

Our primary concern with source rock is its ability to generate petroleum. The amount and type of organic matter present in the source rock dictate not only whether petroleum is generated, but also the amount and type of petroleum.

In a frontier basin we have to answer such basic questions as:

Does the source rock exist?

· Does it have sufficient organic content -- i.e., greater than approximately 0.5%?

· What type of organic matter does it contain?

· How mature is the source?

The type and content of organic matter within a source rock are influenced by the depositional setting and depositional system, which are in turn influenced by basin type and tectonic setting.

High biologic productivity is a key element in the generation of organic material, which may later become source sediments. The highest productivities occur in settings where nutrients and light are abundant. Figure 1 (Abundance of organic matter by common environments.

Figure 1

Upper portion of the figure shows the percent of Earth’s terrestrial or aqueous areas covered by a general environments. Lower portion of the figure shows annual production rate [tons] per unit area [km2]. Total organic carbon produced per year in terrestrial and aqueous areas is approximately equal) shows the abundance of organic matter in common depositional settings.

The type of organic matter produced will also vary according to environment. Non-marine environments other than lakes produce mainly plant debris. Algal material predominates in lacustrine and some marine environments, while planktonic debris predominates in most marine environments. Maturation transforms these organic sedimentary components into petroleum.

It is rare for organic matter to be preserved in sedimentary rocks, because organisms active in newly deposited sediments usually destroy it. Sediment-eating organisms ingest unconsolidated sediments and digest organic material before excreting "cleaned" sediments. Bacteria also digest organic material in unconsolidated sediments. Both types of organic activity require oxygenated conditions within sediments.

Figure 2 (Loss of carbon and related petroleum potential in the sedimentary cycle) illustrates the great inefficiency of organic matter preservation in sediments.

Figure 2

Less than one in fifty thousand carbon atoms involved in life processes will be preserved and accumulate in a petroleum reservoir. More than 80 percent of the carbon will be converted to carbonate. Less than .1 percent of the remaining carbon will be preserved as carbonaceous debris in the subsurface. Of this debris, less than 2 percent will become bitumen. Of this remnant, less than .5 percent will migrate to reservoirs.

In order for organic materials to be preserved, biologic activity within newly deposited sediments must be reduced or halted, generally by rapid deposition or the development of anoxic zones within the water column.

The lacustrine system is the only depositional system within the non-marine setting that produces significant source rocks. Both swamp-derived shales and shales from the deeper lake interior can have high organic content. The organic matter found in swamp-derived shales is vegetation debris, and tends to be preserved because of local reducing conditions. The organic matter in the deep-lake muds is algal and tends to be preserved if the lake is stratified with reducing bottom conditions. Large lake systems with potential source rocks may develop in interior sag and rift basins.

The opportunities for source rock deposition are much more diverse in the marginal marine environment. The deltaic depositional system is particularly favorable for the accumulation of source rocks because of high rates of deposition and high organic productivity. Delta-plain environments consist of marshes and swamps where thick coal layers are deposited in interdistributary shales. Reducing conditions encourage the preservation of organic matter. In pro-delta shales, high organic input is preserved by rapid sedimentation rates, which bury organic materials before they can be consumed. The organic material input into these environments is largely terrestrial vegetation debris. Swamps and marshes also develop in association with barrier island systems, as do lagoons. Lagoons may develop restricted circulation, in which case bottom waters can become reducing and organic matter can be preserved. These depositional systems can develop in almost any tectonic setting or basin type.

Sediments deposited on the shelf can develop into favorable source rocks in areas of high organic productivity, such as zones of upwelling. Source rocks can develop in shallow shelves trapped behind sills, if circulation becomes restricted and reducing bottom conditions are maintained. Sedimentation on the basin plain typically occurs under reducing conditions, and although sedimentation rates may be slow, significant thicknesses of organic-rich sediment can accumulate. Organic matter in these environments will be a mixture of both phytoplankton and zooplankton, with the possible addition of some land-derived vegetation debris. Once again, these depositional systems are not restricted to any particular type of basin or tectonic setting.

In order to evaluate the source potential of a given rock, we must have both quantitative and qualitative measures of organic content.

Maturation

Once we have considered the type and amount of organic content in the source element, our next step is to determine the maturity of the source. We must consider such questions as:

· Is the source immature, mature, or over-mature?

· At what time did the source rock enter the oil window? The gas window?

Potential source rocks exist throughout the geologic column, but most have never been exposed to the temperature and pressure necessary to generate oil. The organic material in sedimentary rocks matures and passes through several stages on

its path to generating hydrocarbons. The first stage, diagenesis, transforms plant and animal organic matter into kerogen, through biological, chemical, and thermal processes. During the second stage, catagenesis, kerogen matures into bitumen, the direct precursor to petroleum. This stage is followed by metagenesis, the intense thermal alteration of kerogen, bitumen, and petroleum into methane and pyrobitumen (solid bitumen). Two important aspects of maturation we must investigate are 1) the organic matter/kerogen type present in the source rock, and 2) the temperature/pressure conditions to which the source has been subjected.

Different kerogen types produce different hydrocarbons during catagenesis. We classify kerogens by the type of organic material from which they are derived

( Figure 1 , Kerogen classification systems.

Figure 1

This summary diagram compares organic matter origin with visual and chemical kerogen classifications and inferred generative character). Type I kerogen, alginite, is usually produced in lacustrine systems and is derived from organic material very high in algal content. During catagenesis, alginite converts into bitumen with efficiencies of up to 80 percent. Type I kerogen is primarily oil-generative. Type II kerogen, exinite, is derived primarily from the remains of plankton deposited in marine systems and during catagenesis generates both oil and gas at efficiencies up to 60 percent. Type III kerogen, vitrinite, is the product of woody vegetation debris and generates small amounts of bitumen during catagenesis. Type III kerogen yields primarily gas, although oil has been generated from some coal units (Thompson et al., 1985; Tissot 1984). Type IV kerogen, inertinite, which is essentially non-generative, consists of recycled or oxidized organic material that can occur in any environment. Figure 2 shows the principal stages of kerogen evolution and the

products generated during catagenesis.

Figure 2

Hydrocarbon maturation is controlled primarily by temperature, which increases within sediments as they are buried to greater depths, or as heat flow within the section increases. Figure 3 shows the relationship between hydrocarbon maturation and depth,

Figure 3

and Figure 4 shows the relationship between maturation and temperature.

Figure 4

In most sedimentary basins, catagenesis begins at depths between 1 and 2 km. Oil generation takes place primarily between 2 and 5 km, the oil window, depending on geothermal gradients and duration of burial. Gas is cogenerated with oil, but continues below the oil floor to depths possibly as great as 7 km in very cool basins, below which source rocks cease generating hydrocarbons. Oil generation begins at temperatures above 60 degrees C, and peaks at a temperature of about 100 degrees C. Condensates form between about 100 and 175 degrees C. Kerogens generate only gas between about 175 and 225 degrees C, and previously formed hydrocarbons will continue to crack into gas up to about 315 degrees C.

There are several physical and chemical characteristics of kerogen we can use in order to determine maturation levels in source rocks within a basin. Physical changes in kerogen during maturation involve a darkening in color, and an increase in vitrinite reflectance. We evaluate both characteristics visually. Chemical characterization involves heating a source rock sample in the laboratory (pyrolysis) to measure the total amount of bitumen present in the sample before heating (called S1) versus the amount of kerogen remaining in the sample after heating (called S2). The relationship between S1 and S2 changes as maturation increases, as shown in Figure 5 .

Figure 5

The temperature at which the maximum S2 response occurs (Tmax) increases as maturation increases.

Source rock deposition and maturation are the first key elements necessary for creating an economic petroleum accumulation. The next key element is migration, which can act to either concentrate petroleum into an economic accumulation, or disperse it and lead to its loss and destruction. Whether concentration or dispersal occurs depends on the presence of porous and permeable carrier/reservoir units, impermeable seals, and trap configurations that can hold a concentration of petroleum in place.

Migration

Most source units are fine-grained, relatively non-porous rocks, such as shales and lime muds. Bitumen generated within source rocks must migrate from these finer-grained units into porous and permeable reservoir rocks before it can be produced. Our concerns about migration include:

· Are carrier units present in close association with the source units?

· Are these carrier units porous and permeable?

· Do migration pathways exist between the source and reservoir units?

Migration occurs in two distinct steps that involve different transport mechanisms. In primary migration, bitumen moves into a permeable reservoir bed, such as a sandstone or limestone. In secondary migration, oil or gas moves through porous and permeable reservoir beds into a trap. Figure 1 illustrates primary and secondary migration.

Figure 1

Primary migration is a familiar topic, but let’s look more closely at secondary migration.

Secondary migration occurs when petroleum moves as discrete oil droplets in a water-wet reservoir unit. Buoyancy and capillary pressure move petroleum through reservoir beds until pore spaces become too small, or until a seal or trap halts further movement. Secondary migration can occur either laterally or vertically. Lateral migration, which occurs predominantly along layers in reservoir beds, can occur over distances as great as 160 km. There are several types of petroleum systems where lateral migration dominates, such as foreland and sag basins ( Figure 2 ,

Figure 2

Figure 3 ,

Figure 3

and Figure 4 , Figure 2-Example of a supercharged, laterally drained, high-impedance petroleum system.

Figure 4

Patterned after the North Slope of Alaska, United States, a foreland basin. Figure 3-Example of a normally charged, laterally drained, low-impedance petroleum system. Patterned after the Williston Basin, United States, a sag basin. Figure 4-Example of a supercharged, laterally drained, low-impedance petroleum system. Patterned after the eastern Venezuela foreland basin. Large accumulations of heavy oil found near the margins of the basin in shallow immature sedimentary strata.). These systems require a laterally continuous regional seal resting on a widespread, permeable reservoir unit, a weak to moderate degree of compressive structural deformation, and uninterrupted homoclinal ramps (Demaison and Huizanga, 1991).

Vertical migration is associated with structural deformation that breaches sealing units through fracturing or faulting. Extensional, wrench, and thrust tectonics produce fault and fracture systems that function as avenues for focused vertical migration, particularly if tectonic activity keeps these avenues open for much of their geologic history (Demaison and Huizanga, 1991). Vertical migration also occurs in pull-apart margins affected by growth faulting and salt tectonics. Rift and pull-apart basin migration styles are illustrated in Figure 5 and Figure 6 (Figure 5-Example of a supercharged,

Figure 5

vertically drained, high-impedance petroleum system.

Figure 6

Patterned after the Central graben, North Sea, United Kingdom. Figure 6-Example of a supercharged, vertically drained, high-impedance petroleum system. Patterned after the Campeche-Reforma Basin, Mexico, a sag basin. The petroliferous part of the basin is vertically drained and has a high impedance [left side of figure]. In contrast, another sector of the basin is laterally drained and has low impedance but is petroleum poor [right side of figure].). Rift basins tend to be vertically drained due to petroleum transfer along faults and fracture systems. Figure 7 ,

Figure 7

Figure 8 and Figure 9 (Figure 7-Example of a normally charged,

Figure 8

vertically drained, high-impedance petroleum system.

Figure 9

Patterned after the Niger delta, Nigeria. Tertiary deltas are usually vertically drained due primarily to the formation of listric faults. Figure 8-Example of a supercharged, vertically drained, high-impedance petroleum system. Patterned after the Los Angeles Basin, United States. Wrench basins typically show vertical migration of petroleum through fault systems. Figure 9- General example of a vertically drained, fold-and-thrust belt. The presence of an effective top seal causes part of the system to show high impedance [right side of figure], while the lack of seals results in a low-impedance sector of the fold-and-thrust belt [left side of figure].) illustrate migration styles in Tertiary deltas, wrench basins, and fold-thrust belts.

Source deposition, maturation, and migration are key elements in petroleum generation and accumulation, but all three are ineffective without a reservoir rock into which the petroleum can migrate

Reservoir

We can often predict reservoir characteristics based on depositional systems models. When evaluating the reservoir element in a play analysis, we must answer such questions as:

· What are the porosity and permeability of our potential reservoirs?

· How thick, laterally extensive and continuous are the reservoir units?

· Are the reservoirs closely associated with the source rocks?

In the case of clastic reservoirs, we can predict that, in non-marine settings, the braided stream system should offer a good environment for the development of

favorable clastic reservoirs. Here, coarse grain size favors high porosity and permeability, although poor sorting may reduce both factors. Braided stream deposits can range from hundreds to thousands of feet in thickness. They tend to be sheet-like deposits ( Figure 1 , Nomenclature of sandstone bodies), with excellent continuity between separate sandstone bodies in the sheet.

Figure 1

In marginal marine environments, excellent reservoir-quality sandstones may occur in several systems, including delta and barrier island systems. Reservoir geometries tend to be elongated rather than sheet-like, so reservoir continuity is poorer here than in non-marine settings. Both sand ridges and sand sheets deposited in the shelf setting may have favorable reservoir geometries, but tend to have poor reservoir characteristics because of low permeability in poorly sorted sands. Slope-fringe sands include slope-apron and submarine fan systems. The former tend to be sand-dominated, while the latter may be sand- or clay-dominated depending on sediment source (Morrill, 1991).

Carbonate rocks with reservoir potential can be deposited in a variety of depositional settings. However, the characteristics in carbonate reservoirs are only partially controlled by depositional processes. In carbonate rocks, diagenesis is often the primary factor in determining which sediments become seals and which become

reservoirs. Carbonate diagenesis can radically alter rock textures, and often determines the ultimate quality of reservoirs.

A basic understanding of carbonate depositional systems allows us to predict potential favorable trends within a carbonate setting, but we must also be familiar with diagenetic controls on carbonate reservoir development (Lloyd, 1987). Diagenesis, however, is a complex topic, involving physical, chemical and biological factors which can interact in countless ways. Also, carbonate rocks often experience several episodes of diagenesis, making it difficult to unravel the true diagenetic history of a carbonate reservoir.

While we may not be able to definitively assess the diagenesis to which a particular carbonate reservoir has been subjected, we can often link diagenetic processes to particular depositional systems and settings. By doing so, we can develop maps of depositional facies, which we can use to make reasonable predictions about the presence or absence of potential reservoir and seal units. These maps may also yield clues about the type and distribution of porosity in the reservoir units.

A majority of petroleum reserves are found in clastic or carbonate reservoirs that retain primary or diagenetic porosity and permeability. Sometimes, however, fine-grained clastic and carbonate rocks can also serve as reservoirs. Fractured limestones, chalks, and shales can be prolific reservoirs with very high production rates, which makes them very attractive targets because of rapid payout for drilling investments. Occasionally, heavily fractured metamorphic or igneous rocks can also serve as reservoirs. Fractured reservoirs are, however, difficult to predict and may be economic only when developed using techniques such as horizontal drilling. The Monterey Formation in southern California and the Austin Chalk in south-central Texas are examples of fractured reservoir plays containing large fields with very high production rates.

Seal & Trap

A seal usually consists of an impermeable unit that overlies or surrounds a reservoir, preventing vertical or lateral movement of reservoired petroleum. Even with ideal source and reservoir units, our play will fail without an effective seal. So, we must ask ourselves:

How laterally extensive are the seals? How thick are they? Do the seals have the optimal structural geometry to be effective?

The simplest type of seal is a contact between the reservoir and an overlying roof rock where this surface has been deformed into a convex-upward shape (Milton and Bertram, 1992). A sandstone/shale couplet deformed into a dome is an example of this type of sealing situation. Other sealing surfaces include sedimentary contacts and facies changes. Typical seals are fine-grained clastics such as shales, fine-grained limestones, or anhydrite and other evaporites. Course-grained rocks cemented with silica, calcite, halite, and asphalt also act as seals. Fault surfaces can also be seals. If a porous unit is juxtaposed with a sealing unit across an inactive fault, the fault generally will act as a seal. Clay or shale distributed in a gouge zone can produce a sealing fault. A pressure difference between two porous units across a fault zone can also create sealing conditions. To truly understand the importance and effect of the seal element, we need to consider its relationship to the trap element.

After petroleum has been generated and has migrated into a reservoir unit, it will continue to migrate through that unit unless it encounters a seal and is trapped in some way. When considering the trap element, we must consider:

Are the traps plausible, given the structural styles we expect to be present in the basin? How are these traps sourced? What sort of mechanisms do we require to source the traps? To seal the traps?

Figure 1 (Nomencalture of a trap using a simple anticline as an example) illustrates the various parts of a classic anticlinal trap, and shows the distribution of oil, gas, and water within a reservoir.

Figure 1

Trap size and geometry also determine the effectiveness of a trap. There are four major types of traps: structural, stratigraphic, hydrodynamic, and combination ( Table 1 ). Trap types CausesStructural TrapsFold Traps Compressional Folds Compactional Folds Diapir Folds

Tectonic processes Depositional/Tectonic processes Tectonic processes

Fault Traps Tectonic processes

Stratigraphic Traps

Depositional morphology or diagenesis (See Table 2 )

Hydrodynamic Traps Water flow

Combination Traps

Combination of two or more of the above processes.

Table 1: Classification of hydrocarbon traps.

Structural traps are created by either folding or faulting. Common fold traps are created by compressional folding due to thrusting or wrenching ( Figure 2 , Northwest-southeast cross section through Painter Reservoir field). Differential compaction may also be a locally important mechanism.

Figure 2

Fault traps require a sealing fault, (see Section 3.5) or an impermeable section across the fault from the reservoir unit. Growth faults create several trapping configurations ( Figure 3 , Schematic cross section of Nigerian field, showing traps and possible accumulation model).

Figure 3

Stratigraphic traps can be classified according to whether they occur within a conformable sequence, or adjacent to an unconformity ( Table 2 ). Figure 4 (a) and Figure 4 (b) show schematic examples of the two classes,

Figure 4

while Figure 4 (c) shows the overlap between classes in the case of channel traps.

Stratigraphic TrapsWithin Normal Comfortable Sequence

Depositional or Facies change Channels Barrier Bars Reefs

Diagenetic Replacement Solution Fracturing

Adjacent to Unconformities

Above Unconformity Onlap Strike Valley Channel

Below Unconformity Truncation

Table 2: Classification of stratigraphic type hydrocarbon traps.

Two subclasses of traps occur within conformable sequences. Depositional or facies-change traps include channels, bars, and reefs, in which porous and permeable units are in depositional contact with seal and trap units. In these traps, depositional geometry creates the trapping configurations. Diagenetic traps are formed when secondary porosity develops in a non-reservoir unit by replacement, solution, or fracturing. Dolomitization of pre-existing limestone creates secondary (replacement) porosity, while solution porosity can develop in both carbonate and sandstone units. Fracturing can affect igneous and metamorphic as well as sedimentary rocks.

Stratigraphic traps adjacent to unconformities are also subdivided into two classes. Those that occur above an unconformity include onlapping sands, and sands that fill channels cut into an unconformity. Those that occur below the unconformity are called truncation traps. In both cases, an overlying shale is the typical seal, and often the source as well.

Hydrodynamic traps rely on downward-moving water within a reservoir unit to restrict upward movement of petroleum. This type of trap is rare (Selley and Morrill, 1983); a combination of hydrodynamic and structural effects is more common.

Combination traps are caused by two or more effective trapping mechanisms. For example, a stratigraphic trap may be enhanced by structural folding or tilting. A truncation at an unconformity may be tilted by faulting to form traps, such as those found in the North Sea ( Figure 5 , Southwest-northeast structural cross section, Piper field, North Sea.).

Figure 5

Diapirism can also lead to several types of traps. These diapir-associated traps include structural, stratigraphic, and combination configurations ( Figure 6 , Schematic of the distribution of potential reservoirs in carbonates [right side of figure] and siliciclastics [left side of figure] during the three stages of salt structure growth.

Figure 6

See original text for detailed explanation of specific trap types.). Both salt and mud diapirs have led to significant hydrocarbon accumulations.

Baisic Geology Ieoc All

Documents

baltimore

los angeles

big horn basin

rocky mountain

lower cook

canadian rocky

western overthrust

upper cook