Structural styles and tectonic history

Deformation never creates a single structure. Extension does not produce one lone normal fault. It produces a system of normal faults with tilted blocks, half-grabens, transfer zones (where one fault system hands over to another) and folds in the hanging walls. Compression produces a system of thrusts and folds. Strike-slip movement produces slanting chains of folds and faults. Such recurring combinations are called structural styles, or structural families. We can recognise the style from a few observations, such as one or two seismic lines, several wells or a geological map. Then we can predict which structures must exist where there are no data, and how they are linked to each other. There is no single agreed list of styles, but everyone recognises the main ones.

Three conditions set the style. The first is which of the principal stresses is vertical. It decides the prevailing type of faults (Anderson's model). The second is whether the basement takes part in the deformation. If the sedimentary cover slides along a weak layer over an almost undeformed basement, we call it thin-skinned deformation. If the faults cut the basement too, we call it thick-skinned deformation. The third is whether the section has weak layers: salt, or shales with abnormally high pore pressure. Detachments (surfaces along which the upper rocks slide) form along them. The rock sequence above a detachment deforms in one way, and the sequence below it in another, so the structures of one tier do not continue into the other.

The style is related to the type of basin (basin types). Rifts are stretched. Foreland basins (basins in front of a mountain range) are crumpled from the side of the mountains. Strike-slip basins develop in the strike-slip style. On passive margins (continental edges that are not plate boundaries), sediments glide down the slope and squeeze out salt. But the relationship is not clear-cut. A basin that began with extension may be compressed later, and its normal faults come back to life with the opposite sense of displacement. One style is laid on top of another. So we determine the style from the observed structures, such as the type and direction of the displacements and the relationships between thicknesses. The type of basin is only a hint about where to look for analogues.

When the crust is stretched, normal faults break its upper, brittle part into blocks. There are two extreme geometries (end-members). Real basins show both, and also the transitions between them.

Domino faults (rotational planar faults)

Rigid blocks are bounded by parallel normal faults that dip the same way. The faults are planar: straight, not curved. As extension grows, the blocks rotate together with the faults, like falling books on a shelf. The faults flatten, and the beds in the blocks tilt in the direction opposite to the dip of the faults. One edge of a block is the footwall (the block below the fault) of the neighbouring fault. That edge is raised and forms the crest of the tilted block. The opposite edge is in the hanging wall (the block above the fault) of its own fault. It is lowered, and a half-graben depression forms above it.

Listric fault

The fault surface is curved. It is steep near the surface, then flattens with depth and merges into a detachment. The hanging wall slides down the curved surface and sags towards the fault, and a rollover anticline forms (rollover; folds and faults). To sag, the wall is stretched. Opposing (antithetic) normal faults and small grabens appear at the crest of the anticline.

Detachment faults and metamorphic core complexes

In very strong extension, the upper crust slides along low-angle detachment faults, and rocks from the deep parts of the crust rise from beneath it. The domes of deep rock that heat and pressure have changed (metamorphic rock) are called metamorphic core complexes. Such structures are rare in petroleum basins, but they show where the stretching “goes” in strongly stretched rifts.

A real rift consists of half-grabens tens of kilometres long (fault groups). The main fault of a half-graben dips one way, and along the rift the direction may change. Neighbouring half-grabens are linked by transfer zones (also called accommodation zones), where the displacement passes from one fault to another (transfer zones). Syn-rift deposits fill the half-grabens as wedges that thicken towards the main fault. The dip of the layers in the wedge decreases upwards, because the lower layers rotated together with the block for longer. The crests of the blocks beside the main faults rise and may be eroded. They then supply coarse clastic material (broken fragments of rock) to the neighbouring depressions. After the extension, post-rift sequences lie undisturbed over the whole system (syn-rift and post-rift stages).

The traps of rifts are: the raised crests of tilted blocks and horsts; rollover anticlines in the hanging walls of listric faults; and syn-rift sandstones that pinch out on the slopes of half-grabens. The source rocks are the deep-water lake and marine shales of the half-grabens, and the seals are post-rift shales (petroleum systems of basins). The main risks are these. Does the fault that bounds the crest seal (fault seal)? Has erosion removed the crest together with the reservoir? Into how many blocks do the faults divide the accumulation?

An example from Kazakhstan is the South Turgai Basin (basins of Kazakhstan). In the Early to Middle Jurassic a rift of alternating grabens and horsts formed here. From west to east they are the Aryskum, Akshabulak, Sarylan and Bozingen grabens, separated by horsts. People link the start of the rift with right-lateral strike-slip movement on the Karatau–Talas–Fergana fault zone. Seismic sections show the flower structures typical of strike-slip faults, and in plan view the faults are arranged en echelon. In the Late Jurassic and the Late Cretaceous the basin was inverted twice. Oil and gas accumulations are concentrated mainly on horsts, on the slopes of grabens and on both sides of the strike-slip zone. So one basin shows three styles at once: extension, strike-slip and inversion.

A fold-and-thrust belt is a zone along the edge of a mountain range. It is tens to hundreds of kilometres wide and hundreds to thousands of kilometres long. In it the sedimentary sequence has been shortened by tens of per cent: it is folded and cut by thrusts. Most often only the cover is deformed. The thrusts detach from a weak layer (salt, clays or mudstones) and run along it on a detachment surface (décollement). Then they cut up through the sequence on a ramp and run along a bed again (ramps and flats). The basement beneath the belt dips towards the mountains and is only weakly deformed.

Thrusts grow one after another. A new thrust forms ahead of the front of the belt, in the footwall of the previous one, and carries the older sheets on its back. So deformation generally gets younger towards the foreland basin (foreland basins). Closely spaced thrusts with the same direction, which meet at depth on a common detachment, form an imbricate fan (a stack of thrust sheets like overlapping roof tiles). If common surfaces bound the sheets above and below (the roof thrust and the floor thrust), and each sheet is thrust over its neighbour, the structure is a duplex. Folds grow above the bends and tips of thrusts (fault-related folds).

Why are some belts narrow and steep, and others wide and gentle? A belt behaves like a wedge of crumpled rock that is pushed up in front of a moving buttress (a rigid block that pushes), like snow in front of a shovel. The wedge grows until its slope reaches a critical value. Then it advances and adds new sheets. The weaker the base of the wedge, the smaller the critical slope. Above salt or shales with abnormally high pore pressure, the belt turns out wide and gentle, with symmetric folds and thrusts that often dip in both directions. Above a strong base, it is narrow, with a steep front and thrusts that dip in one direction.

When the basement takes part in the deformation, reverse faults go down into the basement. Often they are reactivated normal faults of former rifts (see the section on inversion). Above raised basement blocks, the cover is crumpled into steep, asymmetric folds. On platforms (stable parts of the continents), movements on steep basement faults produce gentle drape folds that die out upwards (drape folds).

The traps of fold-and-thrust belts are: anticlines above thrusts and in the hanging walls of reverse faults; traps beneath thrusts (subthrust traps); and reservoir sheets sealed by thrusts. The structures continue along strike for tens of kilometres, which makes exploration easier. But at depth they are complex. Beneath steep limbs and thrusts the seismic image is poor, the crest shifts with depth (crest and hinge), and pore pressure is often abnormally high (Abnormal formation pressures: origin, prediction and recognition). Folds often grow late, and we compare the time of their growth with the time of oil generation (see the section on charge timing).

Fold belts of different ages surround Kazakhstan. The Urals and the Mugalzhar Hills are the suture (the seam) of a Palaeozoic collision of continents, and the Tien Shan is still being compressed today (plate boundaries). In the Permian–Triassic, compression from the side of the Urals formed the subsalt anticlines of the eastern margin of the Pre-Caspian Basin, such as Zhanazhol (salt of the Pre-Caspian Basin).

In strike-slip movement the blocks slide past each other along a vertical fault. In the zone around the fault, the rocks undergo simple shear (a sideways smearing), especially in the cover above a basement fault. In such a zone, shortening and lengthening act at the same time, at 45° to the fault. So structures of all types grow along a strike-slip fault. Folds and thrusts form across the shortening. Normal faults and joints (cracks that open without sliding) form across the lengthening. There are also shears at a small angle and at a large angle to the fault itself: these are Riedel shears. The synthetic R shears have a displacement of the same sense as the strike-slip fault, and the antithetic R′ shears have the opposite sense. Together they form a slanting, en echelon pattern. This pattern shows both that there is a strike-slip fault and which way it moved.

Large strike-slip faults are not straight. At bends and steps where the movement pulls the sides of the fault apart (releasing bends), depressions with normal faults form, up to pull-apart basins. Where the movement pushes the sides together (restraining bends), uplifts with reverse faults and folds form (strike-slip basins). For a right-lateral fault, a step to the right extends and a step to the left compresses. For a left-lateral fault it is the other way round. If the blocks also move towards each other, the result is transpression. If they move apart, it is transtension. In cross-section, a steep strike-slip fault splits near the surface into a fan of faults that looks like a flower: a flower structure. In transpression the faults of the fan are reverse faults, and they push a wedge up. This is a positive (or palm-tree) flower structure. In transtension the faults are normal, and they drop the wedge. This is a negative flower structure.

For exploration, the strike-slip style is the most treacherous. The sign of the vertical displacement and the type of folds change along strike over short distances, from an uplift at a restraining bend to a depression at a releasing bend. The faults divide structures into blocks with different contacts, and a single seismic line shows only a random cross-section of the en echelon structures. Two kinds of evidence prove a strike-slip fault reliably: the sideways displacement of the same features in plan view (river channels, the edge of a platform, the axis of a fold), or slickenside striations (scratches left by sliding) in core. The flower pattern on a single line is only a hint, because other faults crossed at an angle give a similar picture.

Inversion (positive inversion) is the compression of a basin that was formed by extension. The normal faults that bounded the depressions come to life as reverse faults, and the sequence that built up in the hanging walls is partly or wholly squeezed upwards. The opposite case, extension on former thrusts, is called negative inversion. What follows is about positive inversion.

Compression happens most easily on a ready-made weak surface. The hanging wall of a former normal fault, where the thickest syn-rift sequence lies, rises, and an anticline grows above the former depocentre (the place where the most sediment collected). Its core consists of syn-rift layers that thicken towards the fault. If compression was going on while the post-rift layers were being deposited, they thin towards the crest. If the old fault is too steep to slide on under compression, a new, gentler reverse fault appears in the footwall and cuts off the corner (a footwall shortcut). The step of the footwall then acts as a buttress, and the sequence is crumpled in front of it.

In inversion, the displacement from compression is subtracted from the displacement from extension. The deep horizons of the hanging wall were lowered the most, and after moderate compression they are still lowered. On them the fault remains a normal fault. The horizons deposited after the extension are displaced only by the compression, and on them the fault looks like a reverse fault. Between them lies the null point: the horizon that has returned to its former level, where the net displacement is zero. A well that crosses such a fault below the null point will pass a missing interval, and above it a repeated one (missing and repeated section).

The inversion ratio matters for exploration. Weak and moderate inversion, when the null point stays inside the syn-rift sequence, is the most favourable case. The anticlines are simple and clearly visible on seismic data. They lie directly over the former depocentres, where the source rocks are, and these rocks can charge traps both during and after the compression. Strong inversion lifts the source rocks up out of the oil window and stops generation. It also complicates the migration paths, and early traps are eroded or breached by faults.

Inversion is clearly seen on Mangyshlak (folds of Mangyshlak). The northernmost of the Late Permian–Triassic rifts was compressed and strongly deformed before the Jurassic. Its axial part became the Mangyshlak fold zone. The crumpled Triassic rocks there are covered by thin Jurassic and Cretaceous deposits. On the neighbouring Zhetybai Step, the Upper Triassic rocks were eroded before the Jurassic was deposited. We date the compression from such relationships: it is younger than the Triassic rocks it crumpled and older than the Jurassic layers that came to lie on them.

Salt is a structural style in its own right (why salt flows, salt structures). It also affects other styles, because it makes an ideal detachment. In compression, the belt above the salt is wide, with detachment folds that die out downwards into the salt. In extension, the blocks above the salt slide and move apart, and the salt flows in between them and rises beneath the grabens. On passive margins and in deltas, sediments slide down the slope over salt or over ductile shales with abnormally high pressure. At the top of the slope, growth faults (faults that move while sediment is being deposited) with rollover anticlines stretch the sequence. At the toe, the same sliding sheet is compressed into thrusts and folds. Extension at the top and compression at the toe are parts of one gravity system, linked by a common detachment.

In the Pre-Caspian Basin the Kungurian salt divides the section into two tiers with different styles. Beneath the salt are carbonate buildups and the compressional folds of the eastern margin. Above the salt are domes, minibasins and the traps beside their walls (subsalt and suprasalt tiers). We cannot extend the structures of the suprasalt tier downwards into the subsalt tier, or the other way round. The tiers connect only where the salt has been squeezed out completely, through salt welds.

On platforms and in intracratonic basins (basins in the stable interior of a continent; intracratonic basins) the deformation is weak. Basement blocks are raised and lowered on steep faults, and the cover drapes over them in gentle folds that die out upwards. The change from a fault in the basement to a bend in the cover usually happens through a ductile layer that flows over the step. Such folds have a small amplitude, but they continue over a large area and serve as traps on many platforms.

A structure does not carry the time of its deformation as a label. We work out the time from the relationships between structures and rock units (principles of stratigraphy).

• Cross-cutting relationships. A fault, a vein (a crack filled with minerals) or an intrusion (a body of igneous rock that pushed into the layers) is younger than everything it cuts, and older than everything that lies on top of it without being displaced.
• Unconformities. An angular unconformity (a surface where tilted layers are covered by flatter ones) above crumpled layers dates the folding. The folding is younger than the youngest crumpled layers and older than the oldest layers above the unconformity (unconformities).
• Growth strata. Layers deposited while a structure was growing thicken towards the downthrown side of a normal fault and into a syncline. They thin towards the crest of a growing fold, and their dip decreases upwards in a fan. Units deposited before and after the deformation have a constant thickness.
• Structures laid over one another. Folds that have been folded again, faults displaced by younger faults, and joints that stop against older ones (Fractures and naturally fractured reservoirs) put the events in order.
• Numerical age. Biostratigraphy (dating by fossils) and the dating of ash beds give the ages of layers (relative and numerical age). Minerals that filled fractures and faults, and the traces of the cooling of rocks as they were uplifted and eroded, date the deformation itself.

For normal faults that grow during sedimentation (growth faults; more detail), we calculate the history from the thicknesses of growth strata. A unit deposited while the fault was moving is thicker in the hanging wall than in the footwall. The difference between its two thicknesses is the increase in vertical throw during its accumulation. The ratio of the thicknesses is called the expansion index.

A trap will fill only if it existed when hydrocarbons were passing through it, and if it survived afterwards (timing of trap formation). Structural analysis supplies three dates for this check: when the closure arose; when its shape changed (by tilting, growth or erosion); and when the faults that bound the trap last moved.

• A trap older than generation is favourable. We still have to check that late tilting or the growth of neighbouring structures has not changed its shape, and that the spill point has not moved.
• A trap younger than the main generation can fill in two cases. In the first, the oil had been held somewhere before, in earlier accumulations or in the source rocks themselves, and late deformation opened a path for it. In the second, generation was still going on in the deepest parts of the basin.
• The faults of the trap were reactivated after charge. They could have become leakage channels, or they could have filled traps higher in the section.

South Mangyshlak illustrates the second case. According to published data, oil generation there probably began while thick Cretaceous sequences were accumulating, and it continued until the Early Miocene. Uplift and erosion before the Middle Miocene almost stopped it. Meanwhile about 70% of the growth of the anticlines of the Zhetybai Step took place from the Middle Miocene to the present day. According to the interpretation, the accumulations formed mainly in this late stage. Migration into the traps came later than generation, and it could have begun when faults and fractures cut through the mature, overpressured source rocks and the layers above them. A late trap is not necessarily empty, but we must justify such a scenario, not assume it.