Salt tectonics: diapirs, walls, canopies, and welds

Why salt moves

Halite (rock salt) has two properties that combine uniquely:

• Low density: $\rho \approx 2.16$ g/cm³. Freshly deposited mud and sand are lighter than that; clastics become denser than salt only once compaction has buried them beyond about 1 km, so salt is buoyant only below that depth.
• Ductile behavior: salt deforms by viscous flow at very low strain rates. Over geological time (millions of years), salt flows like a very slow liquid while the surrounding sediments behave as brittle solids.

Below that depth the stack is gravitationally unstable, but buoyancy alone rarely lets salt break through a strong roof. What sets it moving is differential loading: sediment piled unevenly, by a prograding delta or a tilted fault block, squeezes salt from beneath the heavy load toward where the load is light. Regional extension thins and breaks the roof so that salt can rise into the gap (reactive, then active, diapirism), and a diapir that reaches the sea floor keeps growing passively by downbuilding: its crest stays near the surface while the sediment around it subsides and thickens (Jackson and Hudec, 2017).

The result: salt migrates through overburden in distinctive ways, producing the geometries we catalog below.

Reading the four geometries

• Diapir. A stem rising from the source layer, here with a bulb overhanging it. Beds drape its crest, are dragged up against its flanks, and sag into rim synclines a little way out, above the source layer that thinned as its salt flowed into the stem. At the figure’s default, $V_{\mathrm{sed}}$ = 2700 m/s, 3.6 km of salt makes the flat target bed at 5.3 km arrive 1067 ms early: read as depth at the sediment velocity, it shows a high 1440 m tall where the bed is flat.
• Wall. Cut across its strike, a salt wall looks like a diapir without the bulb: vertical-sided, tall and narrow. Only a map or a 3D volume shows that it runs on for kilometres. Three-way traps against its flanks are the main targets.
• Canopy. Salt that rose through a feeder and spread at shallow depth as a sheet over sediments younger than itself. The target beneath it is subsalt, not pre-salt: pre-salt means older than the autochthonous salt layer, as in Brazil. The sheet’s gentle top lets vertical rays in, so the target beneath it is lifted by up to 317 ms but lit (beside the feeder, where a column also crosses the feeder’s flared neck, by 416 ms); it goes dark only beneath the sheet’s steep edges.
• Welds and minibasins. Minibasins sink into the source layer until they ground on its base, a primary weld; between them a diapir squeezed shut leaves a secondary weld, often with a teardrop of salt stranded above it. Under the basin centres the salt has gone and the target is not lifted at all; the thick fill above is the record of the salt that left, and the minibasins host turbidite and pinch-out traps in that fill.

Two exercises carry the lesson. Raise $V_{\mathrm{sed}}$ to 4500 m/s and the pull-up and the unlit columns vanish together, though the salt has not moved: the pull-up, $\Delta t = 2h\,(1/V_{\mathrm{sed}} - 1/V_{\mathrm{salt}})$ for salt thickness $h$, measures the salt’s thickness times the slowness contrast, and with no contrast even 3.6 km of salt lifts nothing. Lower it to 2000 m/s and the critical angle falls from 37° to 26°, so more of the bulb’s steep shoulders turn vertical rays away, and the unlit width of the target grows from 0.30 to 0.55 km.

The subsalt imaging problem

Salt presents a unique seismic-imaging challenge. Halite has:

• High P-wave velocity, $V_{\mathrm{salt}} \approx 4500$ m/s, faster than the sediments around it, so every bed beneath salt is pulled up in time.
• Almost no internal reflectivity, and almost no absorption: halite is nearly lossless, so a wave loses amplitude at the salt’s boundaries (a factor $1 - R^2$ for each boundary crossed down and back), not inside it.
• Sharp velocity contrasts at its boundaries: the shallow salt-over-sediment boundary produces strong primary reflections but also strong multiples.
• Ray-bending at the velocity contrasts: seismic energy is refracted as it enters and exits the salt, making correct velocity modeling critical.

The combined effect: beneath a salt body, the image is degraded by several mechanisms at once: rays bent off course, or turned away entirely where top salt dips beyond the critical angle $\theta_{\mathrm c} = \arcsin(V_{\mathrm{sed}}/V_{\mathrm{salt}})$, multiples, and any error in the salt body’s shape in the velocity model. In the worst cases (thick canopies over deep subsalt targets), the subsalt image can be almost unusable without special processing. The technologies that opened up the subsalt plays, namely wide-azimuth acquisition, Ocean-Bottom Node (OBN) surveys, Full-Waveform Inversion (FWI) and Reverse Time Migration (RTM), were designed to compensate for salt-induced imaging distortions.

Salt-related traps

Salt’s unique geometry creates trap styles not found in non-salt basins:

• Supra-salt drape fold: four-way structural closure in sediments directly above a salt diapir or wall. The salt rises, deforming the overlying beds upward into a dome. Very common in Gulf of Mexico shelf fields and some Persian Gulf fields.
• Flank three-way trap: reservoir sand juxtaposed against the lateral side of a salt body, with the salt acting as a lateral seal (halite is almost impermeable). The reservoir rises toward the salt, so hydrocarbons migrate updip and are trapped against the salt flank. Hallmark of Gulf of Mexico subsalt and many salt-wall plays.
• Traps against a weld: where salt has evacuated along a weld, reservoirs can pinch out or dip into the welded structure. A weld seals only where some salt remains; a complete weld often leaks, and welds are common pathways for hydrocarbons migrating up from beneath the salt.
• Subsalt and pre-salt plays: traps beneath salt, sealed by the salt itself. Subsalt targets lie beneath allochthonous salt, a canopy or sheet, and are younger than it: the Gulf of Mexico has produced dozens of major subsalt discoveries since the 1990s. Pre-salt targets lie beneath the autochthonous layer and are older than it: the Aptian carbonates of Brazil (Tupí, Carioca, Júpiter) are the best-known examples.
• Minibasin stratigraphic trap: turbidite sands or stratigraphic pinchouts within a salt-withdrawal minibasin. Classic in the central Gulf of Mexico slope, where minibasins host some of the most prolific fields.

A single salt-province reservoir often combines multiple trap styles: a fault cutting a supra-salt drape, for instance, creates a three-way against the fault as well as the original four-way drape closure. Understanding which trap style applies requires recognizing the salt geometry first.

The global salt provinces

Major salt-tectonic provinces around the world (with the age of the salt):

• Louann Salt (Middle Jurassic), Gulf of Mexico: one of the world’s largest salt provinces. Louann salt was deposited in the early Mesozoic rift phase and has remobilized repeatedly since. The modern Gulf of Mexico is a palimpsest of salt pillows, diapirs, walls, canopies, welds, and minibasins at all scales. Single fields: Lucius, Kaskida, Anchor and Thunder Horse (subsalt); Mars (a deepwater minibasin); dozens of others.
• Zechstein Group (Late Permian), North Sea: the classic European salt province. Zechstein salt walls and diapirs deform the overlying Mesozoic and Tertiary sections. Many producing fields are structured by Zechstein movement. The F3 teaching subset used in Parts 1 and 6 sits above a Zechstein salt sequence at depth.
• Aptian Salt (Early Cretaceous), Santos / Campos Basins, Brazil: the "pre-salt" giants. Tupí (Lula), Carioca, Júpiter, and dozens more sit beneath an Aptian salt layer deposited over Cretaceous carbonates and clastics. The salt acts as a regional seal; the pre-salt carbonates are the reservoirs. Transformative plays that opened new basins worldwide.
• Hormuz Salt (Neoproterozoic-Cambrian), Persian Gulf: one of the oldest salt provinces known. Hormuz salt has deformed for 500+ Myr, producing the iconic salt islands of the Strait of Hormuz and many producing salt-cored fields in Iran and the UAE.
• Ara Salt (Neoproterozoic), Oman: another ancient salt province; similar age range as Hormuz.
• African margin salts: multiple salt provinces in West African margins (Angola, Gabon, Congo), also Aptian in age and the conjugate pair to the Brazilian pre-salt. Major fields like Kizomba, Girassol, and others.

Salt-interpretation pitfalls

• Velocity pull-up. Salt is faster than the sediments beside it, so every bed beneath it arrives early by $\Delta t = 2h\,(1/V_{\mathrm{sed}} - 1/V_{\mathrm{salt}})$: in Figure 3.5, 1067 ms beneath 3.6 km of salt, a false high 1440 m tall. That holds where the salt is encased in slower clastics; where the rocks around it are faster than salt, as in deep carbonate sections, the salt pushes the beds beneath it down instead. Map structure beneath salt in depth, with the salt in the velocity model.
• Confusing salt with other low-reflectivity zones. Basalt flows, some very tight carbonates, and gassy shales can show low internal reflectivity similar to salt. Distinguish salt by its velocity ($V_{\mathrm{P}} \approx 4500$ m/s), low density, and characteristic structural expression (domes, walls, drape).
• Over-interpreting thin salt welds. A thin salt stringer visible on seismic may be a weld (relict of evacuated salt) or may be a real thin salt layer that simply did not mobilize. Check stratigraphic context: welds typically have thick sediment packages on both sides; thin depositional salts do not.
• Ignoring the 3D of salt bodies. 2D sections across a salt body can look wildly different depending on where the line cuts. A line through a diapir’s crest shows a mushroom; a line just off-axis may show only a narrow pipe; a line tangent to a canopy may show only the edge. Map-view control is essential for salt interpretation.
• Subsalt imaging: garbage in, garbage out. The subsalt image quality depends critically on the velocity model used for migration. A wrong salt-body outline in the velocity model will produce a distorted subsalt image that looks plausible but is not real. Iterate salt body interpretation + velocity model + imaging until the result is stable. This is the "salt model building" workflow of modern subsalt exploration.