Folding mechanisms: drape, fault-bend, fault-propagation, detachment

Fold anatomy, the vocabulary

Every fold has a few standard parts. Learning them lets you describe a fold precisely:

• Hinge: the line of maximum curvature, where beds bend most sharply. On an anticline, the hinge is at the crest; on a syncline, at the trough.
• Limbs: the tilted segments of rock on either side of the hinge. In a symmetric fold the two limbs dip at opposite angles; in an asymmetric fold one limb is steeper than the other.
• Axial plane: the plane that divides the fold into two halves (passes through the hinge line at each depth). For symmetric folds it is vertical; for asymmetric folds it leans toward one side.
• Forelimb vs backlimb: in an asymmetric fold produced by tectonic transport, the forelimb is on the side the rocks moved toward (usually the steeper limb), and the backlimb is the trailing limb.
• Plunge: if the hinge line tilts out of horizontal, the fold is plunging. A plunging anticline on a map shows as a nose-shaped contour pattern rather than concentric ovals.

Shape classifications: concentric (parallel layers, constant bed thickness on limbs), similar (bed thickness increases toward the hinge), chevron (angular hinge), box fold (flat crest with steep limbs), dome/basin (doubly-plunging fold with closure in all directions).

Four folding mechanisms

Any specific fold geometry points to one of a small number of underlying mechanisms. These four cover the overwhelming majority of natural folds:

• Drape folds: passive draping over a buried rigid block: a basement horst, a buried reef or an earlier extensional fault block. The loose sediment beside the block compacts under burial and the block does not, so the beds over it sag on its flanks; each younger bed is laid flat on the relief made so far, so the relief decreases upward and the units thin over the crest. (A block pushed up after the beds were laid makes a forced fold of the same family.)
• Fault-bend folds: the hangingwall of a thrust fault rides over a flat-ramp-flat geometry. Each bend in the fault produces a kink band in the hangingwall strata. Suppe (1983) worked out the geometry that keeps every bed's length and thickness: the axial surfaces bisect the angle between the beds on either side of each kink, so the backlimb dips at the ramp angle $\theta$, while the forelimb, folded from the beds the ramp cut, is steeper and narrower. Every bed that has crossed the ramp stands the ramp's height above its regional. This is the workhorse mechanism for thrust-belt anticlines.
• Fault-propagation folds: a thrust fault that is still growing. The beds ahead of its blind tip fold instead of breaking, into an asymmetric anticline whose forelimb is steepest low in the fold, at and below the level of the blind tip, and flattens upward. The more slowly the tip advances for each metre of slip, the tighter the forelimb, until it overturns. The fault may or may not eventually cut through the fold.
• Detachment folds: shortening is absorbed by thickening of a ductile detachment layer (salt, shale, overpressured mud), with the competent cover buckling above. No through-going fault in the competent cover. Symmetric box folds or chevron folds are typical. Because the cover keeps its length and thickness, the ductile layer must flow in to fill the core, and the area it adds is the shortening times its own thickness: a thin layer forces steep, close limbs, while a thick one keeps gentler limbs and an open, box-shaped crest for longer.

These are the fold mechanisms, the smallest units of thrust-belt geometry. Three of the larger architectures they assemble into, imbricate fans, duplexes and pop-ups, are drawn in Figure 3.1 of Section 3.1 (Style: Thrust), so they are not repeated here: learn the mechanism first, then go see the systems it builds.

In the figure below, pick a mechanism and run its history from flat layer cake to today. Plate (a) draws the section at true scale; (b) and (c) measure on it what an interpreter measures on a seismic section, each limb's dip and each marker's relief, level by level, so the four signatures can be set side by side.

Reading the figure

1. In the default fault-bend fold over a 25° ramp the backlimb dips 25°, the ramp's own dip, but the forelimb dips 33°: the fold leans toward the direction of transport although both limbs sit over one ramp. Plate (c) shows the other half of the signature: every bed stands 500 m above its regional, the ramp's height, at every level.
2. Steepen the ramp to 30° and the forelimb jumps to 60°. Above about 30° no simple-step fault-bend fold can keep its beds' length and thickness, which is why the ramp control stops there; ramps that steep fold by other means, such as fault propagation.
3. Drag the deformation back to 0.3. After 600 m of slip the fold stands only 254 m high: relief grows while the hangingwall climbs the ramp, and once the cutoffs pass its top, at 1180 m of slip, more slip only widens the crest.
4. Switch to the fault-propagation fold. At $P/S$ = 2 the reservoir top's forelimb dips 49° against a 30° backlimb. Plate (b) shows the forelimb steepest low in the fold, already just past vertical at 92° on the 600 m marker, and flattening upward to 27° at 1500 m. Lower $P/S$ to 1 and the lowest markers wrap round the tip and overturn further, 149° at 150 m; raise it to 3 and the tip races ahead and cuts the lower beds, and the steepest forelimb left is 56°.
5. Switch to the detachment fold: the limbs are symmetric and the ductile core thickens beneath the crest. Over 300 m of ductile layer, 600 m of shortening has closed the crest and lifts the beds 539 m on 58° limbs. Thin the layer to 150 m and the limbs steepen to 67°; thicken it to 600 m and they ease to 50°, because the core must hold exactly the shortening times the layer's thickness. A thick layer keeps an open, flat crest longest: at half the history, 600 m of it still holds one 135 m wide on 45° limbs.
6. Switch to the drape. Nothing is faulted or shortened, yet under 1500 m of shale the fill top drapes 142 m over a 400 m block and the shallowest buried marker only 14 m. Relief that dies upward is the drape's signature, and plate (c) is where it shows.
7. Finally, press Restore it above plate (a), or drag the history back to 0, for any fold. This is retro-deformation, the QC every thrust interpretation must pass: a section is admissible only if it restores, with bed lengths and thicknesses conserved, to flat layer cake. The table's last cell is the same check read forward, bed length minus span on the reservoir top: 258 m in the default fault-bend fold, 22 % of the slip on the ramp, used up in the fold. Trishear keeps area but not bed length, so for the fault-propagation fold the cell reports the slip itself, all of it taken up in the fold ahead of the blind tip.

Why fold mechanism determines trap style

Each mechanism produces a distinct trap geometry, and the trap geometry determines where hydrocarbons accumulate and how the reservoir is shaped:

• Drape folds: four-way structural closures directly above the buried block. The block itself may be fractured and serve as a deeper secondary reservoir. Classic plays: Arabian Gulf basement-drape anticlines, some North Sea chalk drape traps.
• Fault-bend folds: the ramp anticline provides four-way closure; the ramp fault itself may provide a sealing lateral boundary (via SGR from Section 3.2). Production in these traps comes from the anticline crest, with reservoir units folded up into the trap by the fault-bend geometry. Classic plays: the Alberta Foothills (Turner Valley) and the sub-Andean belt of Bolivia and Argentina.
• Fault-propagation folds: the anticline above the blind thrust is the primary target; forelimb fracturing can produce fracture-enhanced permeability in otherwise tight reservoirs. The blind thrust itself may provide a lateral seal. Common in younger, less mature thrust belts (e.g., Himalayan foreland) where faults haven't yet propagated to the surface.
• Detachment folds: box-fold anticlines with broad flat crests where the ductile layer is thick; over a thin one the limbs steepen and the crest narrows. Very extensive along strike because the detachment distributes shortening broadly. Traps can contain enormous volumes. Classic plays: Zagros salt-cored simple fold belt, Sulaiman Range (Pakistan), Monterrey Salient (Mexico).

Reading the fold mechanism from seismic tells you which of these trap styles applies and therefore where to drill, how to design the wells, and what reservoir heterogeneity to expect.

Reading asymmetry as transport direction

In compressional terrain, asymmetric folds are kinematic indicators, the direction the fold verges toward tells you the direction of tectonic transport. The rule is simple:

The forelimb (steeper limb) faces the direction of transport.

In the Rockies, thrust sheets move toward the east; folds have eastern forelimbs. In the Zagros, transport is toward the southwest; forelimbs face SW. Once you see the fold asymmetry, you can read the transport direction even without picking the thrust fault itself.

The corollary: if you see a fold with asymmetry that points the "wrong way" relative to the regional transport direction, you are looking at a back-thrust fold or a complex doubly-vergent structure. Back-thrusts are common in the triangle zones of fold-thrust belts (e.g., Alberta Foothills triangle zone).

Common fold-analysis pitfalls

• Interpreting folds on unmigrated data. On unmigrated data anticlines look broader and gentler than they are, synclines narrower (bow-ties over tight ones), and every limb dip is too gentle. Always interpret folds on properly migrated data (Section 1.4).
• Artifact folding from processing. Poor statics correction, incorrect velocity model, or mis-picked horizon can introduce spurious folding that isn't geological. Cross-check against well control and adjacent lines before committing to a structural interpretation.
• Seeing fault-bend geometry where there is detachment. A fault-bend interpretation requires visible ramp-flat geometry on the seismic. If you don't see a clear ramp but you do see symmetric folding with detachment thickening beneath, the mechanism is detachment, not fault-bend.
• Salt cores masquerading as drape folds. A fold core that contains salt (Zechstein, Hormuz, Louann, etc.) looks on stack like a buried block producing a drape. But salt flows, dissolves and moves over time, and it does not compact, so the trap geometry depends on present-day salt position, not the original depositional geometry. Always check for salt in the fold core.
• Ignoring 3D. 2D sections under-represent fold plunge and lateral variability. A fold that appears simple on one line may plunge out or complicate on adjacent lines. Always look at along-strike sections and map views of horizon picks.