Structural styles: extension, compression, strike-slip, salt, gravity

Anderson’s fault theory: the framework

In 1905 E. M. Anderson published the framework that still governs how interpreters think about faults. At any point in the crust there are three mutually perpendicular principal stresses: $\sigma_1$, the greatest compression, $\sigma_2$, the intermediate, and $\sigma_3$, the least. A fault does not form on the plane of greatest shear stress, which lies at 45° to $\sigma_1$. It forms where the shear stress first overcomes friction, $\tau = \mu\sigma_n'$, and that happens on the planes at $\beta = 45^\circ - \varphi/2$ from $\sigma_1$, where $\varphi = \arctan\mu$ is the friction angle. For Byerlee’s $\mu$ = 0.6, $\beta$ is 29.5°: the familiar 30°.

Anderson’s second insight is that the ground surface carries no shear stress, so near it one principal stress is vertical. Which one it is sets the fault:

• $\sigma_1$ vertical. The overburden is the greatest compression and $\sigma_3$ is horizontal. Faults dip at $90^\circ - \beta$, about 60°, and are normal faults: the hanging wall moves down relative to the footwall. This is the extensional regime.
• $\sigma_3$ vertical. Horizontal compression exceeds the overburden. Faults dip at $\beta$, about 30°, and are thrusts (low-angle reverse faults): the hanging wall moves up. This is the compressional regime.
• $\sigma_2$ vertical. Both $\sigma_1$ and $\sigma_3$ lie in the horizontal plane. Faults are vertical, strike at $\beta$ to the greatest horizontal stress, and slip horizontally. This is the strike-slip regime.

In borehole and geomechanics work the same three stresses go by their engineering names: the vertical stress $\sigma_V$ (the overburden), the maximum horizontal stress $\sigma_{H\mathrm{max}}$ and the minimum horizontal stress $\sigma_{h\mathrm{min}}$. Saying which of them is $\sigma_1$, $\sigma_2$ and $\sigma_3$ is what connects Anderson to a stress log:

• $\sigma_V > \sigma_{H\mathrm{max}} > \sigma_{h\mathrm{min}}$ in the extensional (normal) regime: the overburden is $\sigma_1$.
• $\sigma_{H\mathrm{max}} > \sigma_{h\mathrm{min}} > \sigma_V$ in the compressional (thrust) regime: the overburden is $\sigma_3$, and it is $\sigma_{H\mathrm{max}}$ that drives the thrusting.
• $\sigma_{H\mathrm{max}} > \sigma_V > \sigma_{h\mathrm{min}}$ in the strike-slip regime: the overburden is the intermediate stress, and both extremes lie in the horizontal plane.

Anderson’s three regimes cover most crustal faulting. Two more styles complete the palette, salt tectonics and gravity-driven listric faulting. Neither is plain Andersonian faulting, but both leave signatures that every interpreter must recognise.

Reading the five styles

The figure opens on a normal fault at Byerlee’s friction. With $\sigma_1$ vertical and $\mu$ = 0.60 the fault dips 60.5°, and 0.72 km of slip drops the hanging wall 0.63 km. From there, each style has its own signature:

• Normal. Planar faults dipping 50° to 70°, beds offset with the hanging wall down, and a package of growth strata, thicker across the fault by the throw, where sediment filled the space the slip opened. A planar fault moves its hanging wall as a block, so the beds stay flat: the figure’s steepest bed dip reads 0°. On a horizon map, parallel faults strike along $\sigma_{H\mathrm{max}}$ with ticks on the downthrown side, and each opens a gap in the horizon as wide as its heave. The Viking Graben of the northern North Sea, the East African Rift and the Basin and Range are the classic examples.
• Thrust. Make the overburden the least stress and the same rock breaks on ramps dipping 29.5°, but only once $\sigma_{H\mathrm{max}}$ reaches 147 MPa at 3 km, more than twice the overburden. Carried up a ramp, the hanging wall folds into an anticline and puts older beds on younger: the repeated section that marks every thrust sheet. Thrust systems build larger architectures from the same unit: an imbricate fan of emergent thrusts off one decollement, a duplex of horses stacked between a floor thrust and a roof thrust, and a pop-up lifted between a forethrust and a backthrust. On the map, thrust traces carry teeth on the hanging wall and strike across $\sigma_{H\mathrm{max}}$. The Zagros, the Andean foothills and the Canadian Rockies are thrust belts.
• Strike-slip. With $\sigma_2$ vertical the fault is vertical and slips horizontally, along its strike and so out of the plane of any section drawn across it. On the section the fault shows itself only indirectly, as a positive flower: strands splaying upward from a vertical root, raised where the fault also converges a little (transpression; a releasing bend or step gives a sagging negative flower). On the map it is plain: markers offset along the fault, and Riedel shears in en-echelon steps at $\varphi/2$ to it. The San Andreas, North Anatolian and Dead Sea faults are strike-slip.
• Listric, gravity-driven. On a delta or a continental slope the weight of a prograding sediment wedge drives extension above a weak detachment (overpressured shale or salt), with no far-field tectonic stress needed. Near the surface $\sigma_1$ is vertical and the fault starts at about 60°, but it flattens with depth onto the detachment. That curvature is what makes a rollover: columns near the fault must drop farther than those away from it to stay on the fault, so the beds bend back toward it (20° in the figure’s Listric style at $\mu$ = 0.60 and the default stage), and growth strata thicken into the fault because they were laid down while it slipped. The Niger Delta, the Gulf of Mexico shelf and the Nile Delta are the examples.
• Salt. Halite cannot hold a shear stress over geological time, so it never faults; it flows. It is not buoyant at shallow depth either: freshly buried mud is lighter than salt’s 2.16 g/cm³ and only becomes denser below about 1.3 km. Most diapirs rise because the overburden is loaded unevenly (differential loading), helped by extension that thins their roof. The salt in a diapir has to come from somewhere, so the source layer thins around it and the roof sinks into rim synclines and minibasins that thicken toward the withdrawal. Beneath salt the seismic image is poor because the fast salt (about 4.5 km/s) bends rays away and its rugose top scatters them, not because salt absorbs energy: halite attenuates very little. The Gulf of Mexico, the Santos Basin and the Zechstein salt of the North Sea are salt provinces.

Side by side the differences are systematic. Normal faults dip steeply and drop their hanging walls; thrusts dip gently and lift them; strike-slip faults are vertical and hide their slip from a section; listric faults curve and bend their hanging walls into rollovers; salt bodies cut across the layering altogether. Section 3.3 draws the folds that faults make in more detail, and Section 3.5 the shapes salt takes.

The global distribution of structural styles

Each style belongs to particular tectonic settings:

• Extensional: rifts (the East African Rift, the Rhine Graben, the North Sea), passive continental margins (the Gulf of Mexico, West Africa) and back-arc basins.
• Compressional: active mountain belts (the Andes, Himalayas, Rocky Mountains and Alps) and their foreland basins. The Zagros Mountains of Iran, one of the world’s largest hydrocarbon provinces, are compressional in origin.
• Strike-slip: transform plate boundaries (the San Andreas, North Anatolian and Dead Sea faults) and their pull-apart basins. Strike-slip alone rarely makes a large hydrocarbon province; combined with extension (transtension) or compression (transpression) it becomes more productive.
• Salt tectonics: basins whose Neoproterozoic to Cretaceous salt was later buried (the Louann Salt of the Gulf of Mexico, the Zechstein of the North Sea, the Aptian salt of the Santos Basin, the Hormuz Salt of the Persian Gulf). Many of the world’s largest accumulations are salt-related.
• Gravity-driven listric faulting: large prograding deltas and slopes (the Niger, Nile and Mississippi deltas). A large sediment supply, a weak detachment and a seaward slope are what drive it.

A single basin often combines several styles in sequence. In the North Sea, Permian and then Triassic to Jurassic extension formed the graben and the main reservoir-bearing basins, Cretaceous to Palaeogene thermal subsidence deposited the chalk, and movement of the underlying Zechstein salt has continued to deform the section since. A careful structural interpretation identifies which style made which feature at each interval.

Polycyclic basins: multiple structural events overprinted

Real basins rarely preserve only one structural style. Successive tectonic events overprint earlier structures and produce composite signatures:

• Inverted basins: extensional basins that are later compressed. The old normal faults reactivate as reverse faults, and rollover anticlines are squeezed into compressional anticlines. The Sunda Shelf (Indonesia) and many northwest European basins are inverted.
• Salt with later tectonics: a salt province later extended or shortened. Diapirs are squeezed shut into welds, where the salt has gone and the strata on either side touch; rim synclines are sheared; faults cut the salt flanks.
• Rifting, then a passive margin: a basin that rifts, opens, and then sits as a passive margin for hundreds of millions of years. The early rift structures are deeply buried, and later gravity-driven listric faulting affects only the upper cover.

Polycyclic interpretation means peeling back the structural events in reverse order: start with the youngest deformation and work backward in time. Cross-section balancing, which you will meet in Section 3.3, is the quantitative tool that keeps this reverse ordering consistent; the figure above moves its hanging walls with one of the simplest balanced constructions, vertical shear.

Common style-recognition pitfalls

• 2D sections hide strike-slip. A strike-slip fault moves rock along its strike, out of the plane of a section drawn across it, so a single 2D line cannot tell it from a steep normal or reverse fault that happens to cut the line. Confirm strike-slip with lines in several orientations and the map-view pattern: offset markers and en-echelon shears.
• Growth strata or later thickening? True growth sedimentation gives each interval a thickening across the fault equal to the throw that accrued while it was laid down, so the throw decreases upward, step by step, with the expansion of each package. Thickening from later compaction or salt withdrawal can mimic growth but will not show that throw-versus-time relationship.
• Imaging failure looks like structural complexity. Subsalt shadow zones, near-surface distortion and velocity-model errors all make images that look chaotic or faulted. Before committing to structural complexity, check the acquisition and the processing history: the chaos may be an imaging artefact, not geology.
• Polycyclic basins are often misdiagnosed. An interpreter trained on a classic extensional basin may miss the later compression that inverted its rollover anticlines; one trained on a thrust belt may miss the earlier extension that set its structural grain. Study the regional tectonic history before interpreting.