Halokinetic hydrocarbon traps in the GOM and elsewhere

In many parts of the world, such as offshore Brazil, offshore West Africa, the North Sea and the Gulf of Mexico, a significant portion of the association of evaporites with hydrocarbons can be tied to salt’s ability to flow, and by its flow generate large and numerous structural and stratigraphic traps. As in bedded evaporites, emplacement of broad sheets of salt against more porous sediment dramatically improves the efficiency of hydrocarbon drainage, both in terms of the area drained and in the tightness of the reservoir seal (Demaison and Huizinga, 1991; Hindle, 1997).

Advances in 3D depth migration and computing power, still occurring today, have revolutionized subsalt and suprasalt hydrocarbon exploration. Only a decade or so ago, most seismic reflection information in halokinetic provinces was acquired as 2D data; today, most seismic data is acquired as 3D data volumes. A 3D cube provides a dense grid of subsurface information, which can be used to map salt body contacts and shapes more accurately. Older vintage seismic used mathematical models for conventional migration and processing, which assumed that the acoustic properties of subsurface sediments did not make abrupt lateral variations. This assumption breaks down near salt flows; a solution required cheap computing power, more sophisticated depth migration techniques and denser data grids. The positioning of contacts in older vintage data can sometimes be improved using more recent processing algorithms and higher computing power, offering a cheaper but less accurate alternative to reshooting a grid.

By the late 1980s, the notion that salt flow could be a horizontal, as well as a vertical emplacement process, was increasingly accepted by the geological community. This new paradigm, along with increased computer processing speed, power and storage, opened up a whole new set of subsalt or more correctly suballochthon or salt tiered plays in the deep waters of the slope and rise in the circum-Atlantic salt basins and similar terranes elsewhere in the world.

At the same time, deepwater drilling technology was improving rapidly, along with geosteering capabilities and broader MWD logging tool suites. Each year wells were being spudded in progressively deeper water. For example, ultra-deepwater exploration in the Gulf of Mexico, in the decade through to the early millennium, has found more than 3-4 billion BOE in nine significant discoveries, in a hydrocarbon exploration province previously considered mature (Figure).

Across the last three decades, offshore exploration in the offshore Atlantic has expanded into a deepwater realm where much of the trap structuration in the continental slope and rise is a response to salt flow, often creating multiple salt tiers in the stratigraphy. In this region, the deeper water offshore province of Gulf of Mexico and more recently offshore Brazil, have been a focus for the development of improved deepwater drilling and seismic technology. 

Technology advances and salt allochthon-related discoveries moving into deeper water in the Gulf of Mexico. A) Distribution of shelf and deepwater salt plays in the Northern Gulf of Mexico. The rectangle indicates the position of B (in part after Cossey, 2004). B) Selected early discoveries, until the mid 1990s, in the updip slope and rise minibasin province in the eastern Gulf of Mexico (drilling in relatively shallow water, near shelf break. C) Water depth to seafloor for deepwater developed fields across the period 1979-2010, with much of the technology expansion driven by Gulf of Mexico (GOM) developments. This figure shows a technology-driven expansion of offshore drilling by the oil industry into ever greater water depths in salt allochthon provinces. D) The number of GOM discoveries for the years 1975-2003, showing reservoir age and dominant play type. It also shows the substantial increase in discoveries from 1995 onward after the passing of the Royalty Relief Act removed some government tax restrictions to deepwater exploration. There were no discoveries in 1978, 1979 or 1992 (after Richardson et al., 2004; Cossey, 2004). E) Hydrocarbon production from the Gulf of Mexico replotted from BOEM data (<http://www.data.boem.gov> last accessed September 9, 2019).

Evolution of the salt play in the Gulf of Mexico

Much of the currently published exploration paradigm for the various deepwater tiered-salt allochthon plays comes from the Gulf of Mexico. However, similar paradigms for allochthon development are being used to drive exploration in other circum-Atlantic basins located in offshore Brazil and West Africa. As yet, this work is not as extensively documented as the Gulf of Mexico, but it constitutes a rapidly expanding database; accordingly, this section focuses on the Gulf of Mexico. Other regions and all halokinetic play paradigms are discussed in detail in Warren, 2016; Chapters 6 and 10 as well as the relevant literature.

Since 1975 more than 23 billion barrels of oil equivalent have been discovered in the deepwater Gulf of Mexico, while some 9.2 tcf of gas and 2,200 mmbl were produced between 1979 and the end of 2002. Deepening water depths of the various discoveries reflect this time-related technological development (Figure A; Table). There were 232 commercial fields discovered in the Gulf of Mexico during the period from 1975 to the end of 2003 (Cossey, 2004). The mean size of these discoveries is 94.3 mmbl oil equivalent. The two largest deepwater fields found to date in the Gulf of Mexico (GOM) are Mars (750 mmbl oil equivalent) and Thunder Horse (1,000 mmbl oil equivalent and possibly as high as 2,000 to 3,000 mmbl).

Thunder Horse was not discovered until 1999, a quarter-century after the 1975 discovery of Cognac, the first deepwater field (Table). Initial well flow rates in wells in the GOM deepwater province did not exceed 10,000 bbl/day until 1995, some 20 years after the discovery of Cognac. Three years later, individual well rates at Ram-Powell and Troika had exceeded 20,000 bbl/day. Likewise, the salt-cored fold and toe-thrust play has been a successful one in the Gulf of Mexico (Figure B). Significant discoveries in the Mississippi Fan (Atwater-Valley) Fold Belt include Neptune (discovered 1995), Mad Dog (1999), Tahiti (2002), Knotty Head (2005), Genghis Khan (2005) and Big Foot (2005) (Table).

Thunder Horse (1999) is the largest of the known fields, with an estimated mean size of 1 billion barrels of oil equivalent (Cossey, 2004). “First oil” from Thunder Horse came online on June 14, 2008. Since then, Thunder Horse has steadily ramped up its production by bringing on new wells. In March 2009, Thunder Horse produced close to 250,000 barrels per day (40,000 m3/d) oil equivalent in oil and natural gas from seven wells. Plans were in place to add two additional wells in 2009 to develop the north end of the field further. By January 2010 total daily production during 2009 had declined from near 250,000 to 175,000 barrels per day (40,000 to 27,800 m3/d). A similar lower than expected rate of production is also seen in wells in Mad Dog field. Reasons for this decline are still being investigated.

Several early subsalt wells were accidental subsalt penetrations where operators were drilling to test anomalous seismic reflectors and found salt where they expected hydrocarbons. The first subsalt well drilled in the Gulf of Mexico Outer Continental Shelf was Placid Oil Company’s Ship Shoal 366 well in 1983. This well encountered three separate salt intervals, but only 90 metres of subsalt sediment were penetrated. Over the next several years, sporadic subsalt drilling continued. One of the more interesting early subsalt wells was Diamond Shamrock’s 1986 well, South Marsh Island 200, which encountered reservoir-quality sand with a massive 245m net thickness. The potential for world-class reservoir sands beneath salt allochthons became apparent with this well, but the absence of reliable subsalt targeting from existing seismic technology limited drilling activity.

Some of the salt allochthon related discoveries in the deepwater Gulf of Mexico over the past few decades (compiled in part from Graham et al., 2011; Richardson et al., 2004; Dribus et al., 2008; Warren 2016).

Porosity and permeability in what are producing deepwater sands in these subsalt fields can be impressive, even though deposition has occurred on deepwater bottoms, situated well out on the continental slope and rise, sometimes in regions that were more than 100 km from their contemporary shorelines. Sands in the Auger Field (Garden Banks 426) retain porosities of 26% and permeabilities up to 350 md. Pliocene and Pleistocene turbidite sands in Green Canyon 205 Field have reported porosities ranging from 28 to 32% with permeabilities between 400 md and 3 darcies. Connectivity in these deepwater sheet sands and amalgamated sheet and channel sands is high for deepwater turbidite reservoirs worldwide, and recovery efficiencies are in the 40-60% range (Table; Cossey, 2004).

High production potential of ultra-deepwater reservoirs in the Gulf of Mexico deepwater was documented in Troika and Mars discovery wells, where initial flow rates from turbidite sands were as high as 31,000 and 50,000 barrels of oil per day per well, respectively, far surpassing the previous Gulf of Mexico single well record of 12,000 barrels of oil per day at Auger. This compared nicely to 500-1,000 b/d from a good well on the shelf in the Gulf of Mexico. Although these GOM fields do not compare in size (storage) to the Middle East, they can generate Middle Eastern rates of production, at least in the early stages. The surprising discovery in 1994 of how productive these reservoirs were in tests at Shell Oil’s Auger platform entirely changed the company’s cost structure for this deepwater play. However, as mentioned earlier, a number of later deepwater projects (including Thunder Horse and Mad Dog in the Gulf of Mexico) have not produced at the peak rates anticipated, and production has declined much more rapidly than predicted, both as a consequence of the reservoir complexity, strength of the water drive, and technology problems arising from remote, high pressure, high temperature subsurface conditions. Accordingly, estimates of ultimate recoverables have been reduced across these fields (Table).

GOM plays_JKW
Gulf of Mexico salt styles cahnge with distance from shelf edge. A) Distribution of the various salt fairways in the Northern Gulf of Mexico. B) Idealised south to north cross section illustrating typical subsalt and suprasalt structures where the position of the salt defines major play types. Note that the low relief pillow folds in the south are autochthonous and probably define the original southern extent of thick Jurassic salt in the basin.

Regionally, there are three subsalt-associated allochthon trends in the deepwater Gulf of Mexico (Figure B, C): i) Flex trend, ii) Minibasin, and iii) Foldbelt (Cossey, 2004). The term subsalt is used by GOM explorers to describe plays obscured by a salt canopy. The flex trend represents the shallowest region of the three plays (in terms of water depth), and it covers a geographic area encompassing the outermost shelf and uppermost slope.

Geologically, it constitutes the halokinetic zone in front of the Plio-Pleistocene Roho, and is an area where there is no widespread salt canopy. It can be further divided into a “toe thrust” or flex belt and a “primary basin” belt (Figure A). Large volumes of salt have been lost in this region via allochthon evacuation, dissolution and weld creation. In terms of salt allochthon evolution, much of the Flex trend and the Primary Basin region has evolved well past the canopy stage and the area is now a relict thrust foldbelt (Figure B; toe thrust zone).

The minibasin play, as the name implies, encompasses sand deposition associated with the development of suprasalt depopods and minibasins (Figure). It is typically located in somewhat deeper water in the Gulf of Mexico than the toe thrust belt, and its exploration began in 1983 with the discovery of Bullwinkle. The foldbelt trend is located beneath and downdip of the canopy belt and encompasses reservoirs developed by compressional folding at the downslope end of a developing allochthon belt. It constitutes the deepest water region of current areas of exploration activity in the Gulf of Mexico.

Flex trend exploration began in the early 1970s with the discovery of Cognac Field located just beyond the present-day shelf edge where there is a “flex” in the seafloor profile (Cossey, 2004). Some of the more substantial flex discoveries were Lena, Zinc, Pompano and Green Canyon 18 (Table). Most discoveries in the Flex Trend are fields with small reserves, characterised by discontinuous sands with producing wells showing relatively low flow rates. Exploration wells in this trend generally targeted “bright spots” on 2-D seismic. Gemini, for example, showed up as a very strong event on the seismic. Texaco was so confident of its Gemini prospect that, based on a very strong event on the seismic, they stopped drilling and set casing just before entering the pay zones. In fact, this hydrocarbon-indicating anomaly could be seen on non-reprocessed, mid-1980s vintage 2D seismic. 

Evolution of minibasins atop salt allochthons based on the northern Gulf of Mexico. A) Evolution of a minibasin as sediment loads a salt allochthon (after McBride et al., 1998a). B) Types of minibasins (after Koch et al., 1998).

Exploration for minibasin plays typically targets the flanks of intraslope basins where reservoir sands pinch out and form combination structural/stratigraphic traps, or it targets zones of drape over turtle structures. Many minibasins plays are sealed and obscured by allochthonous salt sheets, which can fuse into a salt canopy (Figure; Warren, 2016; Chapter 6).

Schematic representation of salt-related geometries in the deep-water Gulf of Mexico (After Pilcher et al., 2011). (A) Interpretation of top primary basin surface and distinction between primary and secondary basins. Salt is red, primary basins are grey, secondary basins are brown, and paired black dots indicate welds. (B) Classification of the top primary basin surface according to the nature of the surface. This classification allows primary basin bounding features such as feeders, ridges, and bucket welds to be mapped. (C) Schematic salt geometry highlighting primary basin trap types: turtle structure (T), bucket weld (B), salt feeder (F), salt ridge (R), base-of-salt truncation (S), and salt cored fold (C).

Some of the larger mini-basin fields discovered to date are Auger, Mars, Diana, Genesis, Troika and Europa. Ram-Powell field was discovered in 1985 and is a very large, stratigraphic trap, developed in a more unrestricted mini-basin and located below a salt tier. Larger minibasins typically have turtle structures developed in their centres. The first mini-basin fields brought on-line (Bullwinkle in 1989, Auger in 1994) produced at much higher than expected rates, had better than expected aquifer support and needed fewer wells to develop them. Much of the early production from these minibasin fields ended up facility constrained. Suballochthon exploration for minibasin traps in the deepwater offshore Gulf of Mexico was first successful south of New Orleans in the region known as the “Isolated Salt Tablets” province or Florida slope. It is located in the easternmost Mississippi Canyon area in waters 1,000 to 2,000 metres deep (Figure; Shirley, 2000; Hall, 2002). Productive reservoirs there are Pliocene to Miocene unconfined turbidites occurring some 3,000-4,500 metre subseafloor. These are usually sealed beneath allochthonous salt and are considered combination structural/stratigraphic traps. Exxon made the first deepwater subsalt discovery of oil and gas in this region with its Mississippi Canyon 211 well. This prospect, nicknamed “Mickey” because of the resemblance of three allochthon sheets in plan view to the head and ears of “Mickey Mouse” has hydrocarbons below the salt canopy and lies beneath 1310 m (4,300 ft) of water. Reserve estimates for “Mickey” range between 50 to 150 million barrels of oil equivalent, but the discovery was deemed subeconomic and has yet to be put on production. Shallow salt canopies partially cover the Thunder Horse field and Pluto discoveries.

The “Mickey” discovery indicated the prospectivity of subsalt plays in this region and focused exploration efforts into sub-allochthons, both on the shelf and in deeper water. As drilling technology and subsalt imaging improved throughout the later 80s and into the 90s, subsalt exploration moved into progressively deeper water. Drilling now reliably reaches larger potential suballochthon minibasin structures, including turtles located further out in the Gulf and targets requiring the drilling of more than 3000 m of salt (Table; Richardson et al., 2004). Minibasin fields currently contain the largest economic resource base in the deepwater Gulf. The subsalt “Mahogany” Field, operated by Phillips Petroleum, is a minibasin play that was discovered in 1993 by the Ship Shoal 349 well in the Mission Canyon region. This field, which began production in December 1996, was the first commercial subsalt oil development in the Gulf of Mexico. In 1994, Shell Offshore, Pennzoil, and Amerada Hess announced another significant minibasin discovery in Garden Banks 128 well, nicknamed “Enchilada” (as in the vernacular “it’s the big enchilada”). Enchilada has combined reserves estimated at 400 bcf of gas and 25 million bbl of oil/condensate (DeLuca, 1999), and was brought online in July of 1998. In 1996, nine subsalt wells were drilled, and three were discoveries. The largest discovery (Anadarko and Chevron) was “Gemini” in Mississippi Canyon Block 292, with estimated reserves of 250-300 Bcf of gas and 3-4 million bbl of condensate (DeLuca, 1999). It currently produces from a single reservoir using a subsea system. Gemini was the first deepwater subsalt production in the Gulf of Mexico - and showed explorers the potential for prolific output from the deeper water portion of the subsalt play. In June 1999 one of the Gemini wells tested around 76 million cubic feet of gas and 1,370 barrels of condensate a day. Agate Field in Ship Shoal 361 was discovered in 1996 by Anadarko and Phillips and is produced through a tieback into the neighbouring Mahogany platform. Anadarko, along with partner BHP, also announced a second subsalt discovery in 1996 in Vermilion 375, called Monazite. The discovery well revealed multiple pay zones, but because of problems during testing, the hole was plugged and abandoned. Amerada Hess and Oryx made a major discovery in late 1997 in their Penn State prospect in Garden Banks 216; the field came on stream in 1999 through a tieback to the Baldpate production facilities. In July of 1998, Anadarko announced a subsalt discovery at the Tanzanite prospect in 95 m (314 ft) of water in Eugene Island 346 with reserves estimated at 140 million BOE. Later in 1998, Anadarko also announced the discovery of their Hickory prospect in Grand Isle 116 in 97 m (320 ft) of water. The discovery well, drilled to a total depth of 6580 m (21,600 ft), penetrated approximately 2400 m (8,000 feet) of salt before passing into the reservoir. Reserves are estimated at 40 million BOE (DeLuca, 1999).

Thunder Horse Field (formerly known as Crazy Horse Field) in Mississippi Canyon Blocks 776, 777 and 778 is a substantial minibasin field with a turtle style of reservoir development. It is located in the Boarshead Basin, 200 km south-east of New Orleans and is perhaps the largest deepwater discovery made to date in the Gulf of Mexico (Enger and Logan, 2001). It has estimated reserves of greater than a billion barrels of oil equivalent in early and middle Miocene sands at about 7,000 - 8500 m (23,000-28,000 feet) depth. BP-Amoco has said reservoir characteristics are excellent and compared Thunder Horse favourably to Magnus Field in the UK North Sea, where wells flowed at rates exceeding 20,000 bbl/day at peak. Thunder Horse shows that deepwater subsalt sands can be productive, even at depths well below 7500 m. It is one-and-one-half to twice the size of the next biggest structures in the region (Enger and Logan, 2001) and is one of three large discoveries on a turtle structure worldwide. Other turtle-associated discoveries are Pluto (aka BS and T; Mississippi Canyon 718) and Mensa (Mississippi Canyon 731). Both the Thunder Horse field and Pluto discoveries are partially covered by shallow salt canopies.

Methods of most effectively exploring for minibasin plays are still emerging, especially if a turtle is obscured by a salt canopy. Deeper turtle structures are still difficult to reliably image on 3D seismic data, especially beneath thick salt canopies and there are no diagnostic hydrocarbon-indicating amplitudes (bright spots) related to this play, so structures rather than accumulations are targeted. Geologically, turtle structures can be complicated as these structures were synclines at the time they were receiving sediments (Figure 6.29). Then the structure inverted due to ongoing salt withdrawal. That in itself implies there is a complicated relationship between the timing of trap formation, hydrocarbon charge and migration. The critical uncertainty with turtle structures in this style of play is whether a trap was in place during the hydrocarbon migration phase. Also, there can be crestal faulting on these structures that can degrade the top seal of the traps, but the size of the potential discoveries makes the risk worthwhile. Petroleum systems in the foldbelt give prominence to underlying high-relief salt-cored folds (mostly to the west) and thrusted folds to the east in what is called the Atwater Valley style structural province (Hall 2002). Lower and middle Miocene turbidite reservoirs extend across the crests of folds, which typically formed during foot of slope compression in the mid-Miocene to Early Pliocene.

Three major foldbelt trends are considered prospective: the Mississippi Fan fold belt (figure), the Perdido and the Port Isabel foldbelt (figure), all are located in deeper water downdip of the salt canopy/minibasin trend. The Foldbelt play began to be exploited in 1995 with BP-Amoco’s discovery of Neptune in Atwater Valley 575 in the Mississippi Fan foldbelt. Soon after that, BP discovered Atlantis and Mad Dog (Figur;e Hall, 2002). K2/Timon was found just north of the earlier discoveries, and more recently Texaco discovered the Champlain Field in the middle of Atwater Valley, extending the play further east (Table, Rowan et al., 2004). To date, in the Mississippi Fan belt, a significant portion of the discoveries are centred in the foldbelt of the western part in a small area of the southeast Green Canyon area. All are near the edge of the Sigsbee Escarpment and within 50 km of each other in water depths ranging from 1,200 to 3000 metres (4,000 to 6,500 ft).

Section across the Mississippi Fan Fold Belt, based on seismic line in Meyer et al. (2005). Indicates regional late Miocene salt inflation.

 Target reservoirs are found in the 5,000 to 8,000 m (17,000 to 26,000 ft) depth range, and the structures can be prolific with greater than 300 million barrels of oil equivalent. Mad Dog is currently one of the more productive of the foldbelt discoveries, with reserve estimates ranging from 400 to 800 million barrels of oil equivalent from about 100 m of net hydrocarbon pay in a large, high relief fold under the edge of the Sigsbee Escarpment (Figure). Atlantis vies with Mad Dog in terms of field size. Current seismic quality in this part of the fold-belt province is good to fair, but companies are exploring very high relief features, which are difficult to miss. For example, Mad Dog is a 3,000 m relief feature below the Sigsbee Nappe. 

Port Isabel Foldbelt, northwestern Gulf of Mexico. A) Regional depth section showing linked proximal extension and distal contraction. Shortening (compression) is accommodated in: the Perdido Foldbelt above the Jurassic Louann Salt; extrusion of, and translation above, allochthonous salt; and the Port Isabel Foldbelt, detached on Eocene shales (2x1 vertical exaggeration). B) Seismic interpretation of the Port Isabel Foldbelt, showing multiple shale detachment levels and associated fault-bend and fault-propagation folds. C) Schematic cross section (not to scale) showing the relationship of the Port Isabel shale detachments (at the top of the green unit and higher) and the underlying, autochthonous Louann Salt detachment (after Rowan et al., 2004; Morley et al., 2011)

Even though imaging is not the best quality, it’s hard to overlook a 3,000-metre-tall closed structure. Plays in the area known as the Eastern Sigsbee Escarpment do not involve the search for the subtle trap! The province’s western portion is relatively well drilled now, so the play’s future lies in taking the charge risk and moving east to determine how far the play will extend. In contrast, the Perdido Foldbelt Play is underexplored. The play is similar to the Mississippi Fan Foldbelt Play in that it consists of enormous compressional foldbelt structures. Like the Mississippi Fan Foldbelt Play, the compressional folds extend a considerable distance landward beneath the Sigsbee Salt Canopy. However, the plays differ in the age of the structures and the targeted reservoir. Early exploration drilling in the deepwater Gulf of Mexico in the early 2000s found over 2 billion BOE (Richardson et al., 2004).

Today the traditional deepwater mini-basin plays are still providing many successful exploration opportunities (e.g. the Thunder Horse and North Thunder Horse discoveries in southern Mississippi Canyon), but recent developments in new foldbelt deepwater plays continue to expand the exploration potential of the deepwater Gulf of Mexico (e.g. Perdido). Currently, some 99% of total production in the deepwater Gulf of Mexico is from Neogene reservoirs (Pleistocene, Pliocene, and Miocene). Several recently announced deepwater discoveries encountered large potential reservoirs in sands of Paleogene age (Oligocene, Eocene, and Paleocene); for example, the Tiber discovery in 2009 and the Coronado Discovery in March 2013. Transocean’s Deepwater Horizon semisubmersible made the historic discovery for BP at the company’s Tiber prospect in the Keathley Canyon in September 2009. Drilling in 1220 m of water and to a world-record total depth of 10,685m, Deepwater Horizon tapped in a pool of crude estimated to contain 4 to 6 billion barrels of oil equivalent, one of the most significant U.S. discoveries.

Mad_Dog_Rowan 2004_jkw
Depth section through the western part of the Mississippi Fan Foldbelt showing a series of symmetrical, salt-cored detachment folds between welded synclines and the resulting closure that formed the Mad Dog and K2-Timon fields (after Rowan et al., 2004).

Six months later, in March 2010, Shell (with partners Chevron and BP) started production at its Perdido spar in 2,438 m of water in the Alaminos Canyon. A hub for the development of three fields, Perdido was the world’s deepest offshore platform, and the first project to pump oil and gas from the Lower Tertiary in the Gulf of Mexico. In 2010, Petrobras began to develop the Gulf ’s first floating production, offloading, and storage facility to produce from Lower Tertiary reservoirs at its Cascade and Chinook prospects. By end 2010, the industry had announced 19 discoveries in the Lower Tertiary trend, 14 of them containing more 100 million barrels of oil equivalent. However, the aftermath of the BP Horizon spill in April 2010 was a decline in exploration, from which the Gulf is still recovering. Deepwater salt structure-related discoveries continue to be made in the Gulf of Mexico and many other salt basins worldwide (see Warren 2016. Chapters 6 and 10) for a full discussion, along with examples, of the halokinetic exploration paradigms.

© 2016-2024 Saltwork Consultants Pty Ltd