Hydrocarbon Exploration in the Gulf of Mexico and Caribbean, a Plate Tectonic Perspective

Abstract

Since the early 1900s, oil and gas exploration has progressed with great success in the Mesozoic-Cenozoic basins of the Gulf of Mexico (USA, Mexico) and northern South America (Venezuela, Colombia, Trinidad, Guayanas). Much of these margins are classic passive margins, with offshore geology comprising more distal equivalents of the onshore. However, the Caribbean margins of northern Colombia, Venezuela and northern Trinidad were shown early on to be different, with basements of igneous and metamorphic rock lying immediately beneath various Cenozoic deposits. A similar situation was clear by the 1950s in Cuba, as well. The vast majority of the region’s produced and remaining proven reserves occur in the Gulf of Mexico and Proto-Caribbean rifted passive margins. For the Gulf of Mexico, these include the Mesozoic and Cenozoic strata of the northern rifted margin (Burgos to South Florida basins) and southern rifted margin (Salinas Basin to Cozumel) rifted margins, and the narrow, western Gulf transform margin (SE Tamaulipas Arch to Veracruz). Plays are numerous, from Mesozoic shelf carbonate and sand systems, to slope channels and deep-water fans. Traps are numerous as well, with stratigraphic, structural and combined traps from the Oxfordian to the Pleistocene. Although “passive” margins, gravitational motions have certainly played a role in overall margin deformation, such as downdip extensional-contractile detachment systems. In addition, a gentle tectono-magmatic event affected the East Texas-Northwest Louisiana area, forming the combined traps of the super-giant East Texas fields. Further still is the entire discipline of salt-related exploration, from the flanks and tops of salt diapirs, to sub-salt traps beneath detached salt sheets and canopies, to mini-basins and other salt evacuation structures. Finally, there remains the whole of the sub-salt play to be tested, with much of the Eagle Mills and Todos Santos formation sandstones looking like better reservoir targets than mere syn-rift red beds. Indeed, sag sections overlying syn-rift sections are well developed (Pindell et al., 2011; 2014; 2018), so the big question regards the possibility of a sub-salt source rock. All in all, the Gulf of Mexico margins are sufficiently dynamic and laden with ever-changing sedimentation that hydrocarbon maturation has been ongoing, with numerous “critical maturation moments” depending upon location. Moving to the Caribbean, my version of the Pacific origin model for the Caribbean oceanic crust and arcs holds that a swath of oceanic plateau-dominated lithosphere of the Farallon Plate, led by the Cretaceous Caribbean island arcs, migrated into the widening Proto-Caribbean Seaway, subducting the Proto-Caribbean lithosphere as it came. The leading arcs and their oceanic forearcs collided obliquely and diachronously with the Proto-Caribbean passive continental margins, forming the eastward-younging circum-Caribbean suture and foreland basins above the older Proto-Caribbean passive margin sedimentary sections. These include the Maastrichtian “Sepur Basin” in northern Guatemala; the Early and Middle Eocene “Cuban Basin” in the Florida Straits; the Early and Middle Eocene “Misoa-Trujillo Basin” in Maracaibo; the Late Eocene-Oligocene “La Pascua-Roblecito Basin” in central Venezuela; the Late Oligocene-Middle Miocene “Los Jabillos-Carapita Basin” in Eastern Venezuela; and the Late Oligocene-Middle Miocene “Cipero-Naricual-Brasso Basin” in Trinidad (Pindell et al., 1988; Pindell et al., 1991). This eastward-younging progression records an average Caribbean-American displacement rate of about 20 mm/yr since the Maastrichtian. In addition, two intra-arc basins formed within the leading Caribbean arcs during this migration, the Yucat·n and the Grenada/Tobago Trough basins. Both played important roles as the Caribbean Plate accommodated the pre-existing shape of the Proto-Caribbean Seaway, by allowing the arcs to expand under extension until they encountered (collided with) a Proto-Caribbean margin. A third intra-Caribbean basin within the original plate is the Eocene and younger Cayman Trough, which has recorded some 1100 km of displacement between the Caribbean Plate and the Americas over that time. This displacement is also recorded by the Lesser Antilles arc magmatism since the Middle Eocene. Given this simple summary of Caribbean tectonic history, the hydrocarbon habitat of the Gulf of Mexico and Caribbean region can be assessed from a plate tectonic perspective that acknowledges all the mobilistic implications of the many and various geological elements of the region (e.g., Pindell, 1991; 1995). Below I discuss some of the main points made by these papers, as well as new points in light of newer data. Concerning the Proto-Caribbean margins (Bahamas, southern and eastern Yucat.n, and northern South America), a primary structure of great significance is the “circum-Caribbean suture zone”. At the surface, this suture separates the distinctly different Caribbean from Proto-Caribbean suites of rock, with very different exploration potential. One point that should be stressed is that the suture does not necessarily correspond to today’s Caribbean plate boundaries. For example, the arc portions of Central and Eastern Cuba and the entire Yucat.n Basin and Cayman Ridge are no longer part of the migrating Caribbean Plate. Likewise, along much of northern South America, the suture lies onshore, and the famous “La Luna” source rock section is not known in the offshore. Further, the northern South American margin was probably convergent prior to the arrival of the Caribbean Plate (Pindell et al., 1991; Pindell and Kennan, 2007; Pindell companion paper, this meeting). Only after Caribbean collision with South America did the plate boundary shift offshore (Dewey and Pindell, 1985; 1986: Pindell and Erikson, 1994). A second point that should be stressed is that the circum-Caribbean “suture”, both at surface and at depth, is rarely a “hard boundary”. Almost everywhere, Proto-Caribbean strata are imbricated beneath the hanging wall thrust zone of the Caribbean suture, such as in Cuba (Iturralde et al., 2008; Garcia-Casco et al., 2008; Stanek et al., 2009). Further, Caribbean accretionary prism strata were originally deposited on the Proto-Caribbean seafloor, but subsequently were accreted and transported as part of the allochthonous arcs, presenting a semantic issue. Terms like para-allochthonous or para-autochthonous can help, but the user needs to make sure the receiver knows what is intended. Further still, once sutured, rocks from one suite may share roles of rock in the other suite. For example, in Cuba, fractured serpentinites and sandstones in the para-allochthonous Placetas accretionary prism, both in the Caribbean hanging wall, serve as reservoirs within the suture zone due to upward migration. In the Proto-Caribbean margins outside the circum-Caribbean suture, the mid-Cretaceous is thought to have been the time of richest source rock deposition with the greatest volumetric expulsion potential (e.g., Napo, Villeta, La Luna, Querecual, Canje, Cob.n fms). However, the Upper Jurassic of western Cuba’s Sierra Guaniguanico (e.g., Jagua Formation), which represents initial marine sedimentation along the eastern Yucat.n margin (Pindell, 1985a,b), has good source character, and broken samples smell strongly of hydrocarbons. Elsewhere, such as along the Venezuelan and Trinidadian coasts, the Jurassic units presently comprise highly graphitic phyllites suggesting considerable original TOC content. But the Proto-Caribbean Mesozoic section is rarely more than 2-5 km thick. Even if we acknowledge the Upper Jurassic’s source potential, the post-break up marine sedimentary sections do not appear to have been thick enough prior to the initiation of circum-Caribbean foredeep deposition at any particular locality for initiation of thermal HC maturation of source rocks. This means that the majority of hydrocarbon maturation in the Proto-Caribbean margins is directly related to the approach and oblique collision of the Caribbean Plate with those margins. However, there are two ways in which maturation can be achieved; one, burial by structural imbrication of para-autochthonous section and emplacement of Caribbean elements above that, and two, burial by foredeep basin deposition on the down-flexed autochthonous footwalls. In addition, in northern Colombia and western Venezuela, uplift of the northern Andes chains has post-dated the original collision in that area of the Caribbean Plate. Thus, there are Neogene “Andean” foredeep basin sections that have been superimposed upon the earlier Paleogene Caribbean foredeep sections, with correspondingly younger pulses of maturation. The maturation history in Lake Maracaibo Basin is a good example: oils matured in the Eocene due to Caribbean collision there migrated south from under the Caribbean thrust front and thickest parts of the Misoa-Trujillo foredeep section, whereas oils matured in the Miocene due to Andean uplift migrated north/northeast to the Bolivar fields from under the thickest parts of the Andean Guayabo foredeep section. In northern Monagas and the Matur.n Basin, maturation is Miocene and younger because the Caribbean Plate did not arrive there to produce a foredeep section above the passive margin section until the Miocene. The above is not to say that maturation would not have happened along the Proto-Caribbean margins were it not for the migration of the Caribbean Plate. Indeed, the recent discoveries in Guayana where collision never occurred show otherwise. However, along northern South America for example, the largely Paleogene arrival of the Caribbean Plate has interrupted and deflected much of the clastic transfer that would have occurred across the original Proto-Caribbean margin, and thus the collision of the Caribbean Plate and development of its foredeeps, rather than those clastic sections, are responsible for maturation. Turning to the Caribbean realm, we need to consider very different geology: oceanic plateau basement (Colombian and Venezuelan basins), arc systems (all the Antilles and Central America), extensional intra-arc basins (Yucat.n Basin and Grenada Basin-Tobago Trough), accretionary prisms (Muertos Prism, South Caribbean Prism, North Panama Prism, Lesser Antilles-Barbados Prism), the Cayman Trough pull-apart basin, and a myriad of smaller basins formed by crustal faulting and releasing and restraining bends related to large-scale strike slip fault systems. Concerning the crust of the plate interior, an entirely oceanic crustal character continues to be indicated as more and more data are collected. In the Colombian and Venezuelan basins and the lower Nicaraguan Rise, deep-sea drilling and seismic reflection and seismic refraction data indicate an entirely basaltic basement, often of the oceanic plateau type, with crustal thicknesses ranging from 6 to 25 km. The thicker parts of these basaltic complexes are often layered in seismic reflection records, and the layering can be locally similar to that of SDR (seaward-dipping reflector) complexes along magma-rich continental margins. However, the SDR appearance is also seen at other clearly intra-oceanic settings, such as in magmatic build ups in the Indian Ocean and the R.o Grande and Walvis hot-spot complexes. The pattern is not an indicator of proximity to continental margins, but rather of excessive magmatism at sites of basement extension such as magma-rich continental margins and near-axis (spreading ridge) oceanic plateaux (e.g., Iceland, Galapagos). The layered appearance should not be confused with sedimentary layering. In part of the central Colombian Basin, free-air gravity maps show a “broken-up” pattern which may indicate a small area of thin crust with a magma-poor spreading character. Heat flow into the Caribbean stratigraphy should not pose a great risk to exploration, because oceanic crust that is only 80-90 m.y. old should theoretically have similar heat flow to continental crust. Furthermore, Caribbean arc mineralogies, as well as widely distributed arc detritus including tuffs, are fairly potassium rich, and potassium’s decay to argon generates heat. The Caribbean arcs (Greater Antilles, Aves Ridge, Lesser Antilles, Leeward Antilles, and Panama-Costa Rica) are generally of the intra-oceanic island arc type, according to geochemistry, petrology, and geochronology. These magmatic build ups presumably have developed upon and overlie former oceanic crust that is at least as old as the overlying arc. Some of the Caribbean Cretaceous-Cenozoic arc magmatic rocks have recently been shown to contain Proterozoic and Paleozoic zircons. However, the geochemistry of these rocks shows typical island-arc trace element patterns and Strontium ratios. The favoured interpretation is that the zircons were not acquired from crustal contamination as magmas rose through continental crust, but rather that the zircons were sourced from the mantle along with the magmas (e.g., Rojas-Agramonte, et al., 2015). Contamination of the mantle wedge by the addition of sediment during subduction is well established, so it is no surprise that ancient zircons can rise from the wedge along with the island-arc magmas. Concerning source rocks, DSDP drilling (Leg 15) reported thin (cms, hence low volume) Coniacian layers with about 2-3% TOC. These have been re-measured more recently with claims of TOC values as high as 5% or even 8%. These layers most likely relate to Oceanic Anoxic Events (OAE) and may be correlative to other OAE levels in other oceans, as well as to the better known northern South American shelf margin equivalents, the Napo, Villeta, La Luna, Querecual, Naparima Hill and Canje fms. Flexural backstripping can be employed to determine the approximate depth of deposition of the Caribbean Coniacian section. Where the crust is typically oceanic, this will be about 2.5-3 km deep. But in oceanic plateau areas with thicker crust, the depositional setting would have been significantly shallower. Portions of the Caribbean plateau were probably subaerial, much like the Galapagos Islands today. This probably explains why the Coniacian is silty (volcanogenic) in the DSDP boreholes. In Pacific origin models for the Caribbean, the Caribbean crust was situated between Mexico and Colombia in the Coniacian as opposed to directly north of South America. It is not obvious why this paleo-depositional position would be better or worse for source rock quality and volume than a position closer to today’s geography. Tertiary source rocks are also known in the Caribbean, such as the Late Oligocene-Early Miocene shales of the Falc.n Basin and the Gulf of Venezuela, which can interfinger with shallower water [carbonate] correlatives with reservoir potential, all deposited during renewed subsidence after Eocene Caribbean collision (e.g., La Perla discovery). The fact that Caribbean basement lies along the coast of and offshore northern South America, rather than a more distal equivalent of the onshore Cretaceous section, does not mean the Venezuelan offshore should be overlooked. Concerning reservoirs, quartz is rare in the Caribbean, and the predominance of mafic and feldspathic minerals in the arcs, which weather to clays, poses a significant risk to clastic reservoir quality. The exception to this of course is the southern Caribbean boundary with South America, where quartz-rich detritus can reach Caribbean basins (Kolla et al., 1984; Pindell et al., 2009). Beyond clastic reservoirs, buried reef complexes are always possible, as are zones of fractured rock, such as the weathered “granite” at La Vela Bay in Venezuela. In addition to the lack of quartz, the volume of clastic detritus over the Caribbean Plate is generally low, due to the smaller erosional catchment areas (islands rather than continents). Again, the exception to this is along northern South America. But elsewhere, achieving sufficient burial for thermal maturation is a significant risk. Consequently, the areas of contractile imbrication provide an alternative means for achieving thicknesses capable of initiating thermal maturation. These include the Muertos Prism along southern Dominican Republic and Puerto Rico, the South Caribbean Foldbelt, the North Panama Foldbelt, and the Lesser Antilles-Barbados Prism. Concerning the Yucat.n and Grenada-Tobago Trough intra-arc basins, seismic stratigraphy and geophysical observations point to an early Paleogene age for both. Parts of both basins are floored by oceanic crust, so heat flow should be reasonably high (young basins). Nevertheless, the Yucat.n Basin may lack sufficient sedimentary section to have initiated thermal maturation of hydrocarbons. In contrast, the Grenada and Tobago Trough basins, once a single Paleocene-Eocene basin but divided by the younger southern Lesser Antilles arc, have more than 8 km of section, and much of this probably derives from quartzose clastic transfer systems in western Venezuela. Also, these basins were potentially silled from deep water circulation for parts of their histories, beneficial to source rock deposition. The southern Grenada Basin margin is a zone of basement-involved backthrusting resulting from the collision with the Serran.a del Interior Oriental. There are also gravitational slumps in the southern parts of both sub-basins, and backthrusting of the Barbados Prism has produced folds deep in the eastern Tobago Trough section. These two sub-basins have all the main elements for generating commercial hydrocarbon reserves; the recent Nutmeg discovery may point to more to come. Finally, I wish to point out two places in the Caribbean where continental crust possibly underpins Caribbean island-arc crust. The first is the upper Nicaraguan Rise-Jamaica, and the Cayman Ridge. As discussed in a companion talk (Pindell, this meeting), the reconstructed upper Nicaragua Rise-Jamaica, Cayman Ridge and Central Cuba (western end of the Greater Antilles arc) collided with the southern Yucat·n continental margin in the Maastrichtian (Pindell and Dewey, 1982; Rosenfeld, 1993; Ratschbacher et al., 2009). This collision likely occurred on relatively low-angle thrust detachments, such that the arc overrode the distal Yucat·n margin and parts of its sedimentary marginal cover. However, the ensuing strike slip history along the paleo-Motagua fault most certainly occurred on much higher angle (vertical) faults, thus potentially subcreting elements of the distal Yucat·n margin to the base of the obducted arc crust. Whether or not continental crust remains beneath the Jamaican arc, such overthrust sediments may be the source of certain oils reported in or around Jamaica. The second place is the Aves Ridge. Maresch et al. (2009) made a case that Margarita and the Aves Ridge originated from the western margin of Colombia, based on the ages of basement units and the timing of HP-LT metamorphism. The simplest model that explains the data is one in which the Aves Ridge was an active arc separated by intra-arc extension (the Colombian Marginal Seaway; Pindell and Erikson, 1994) from the remnant Colombian (Central Cordillera) margin in the Late Jurassic and Early Cretaceous. Usually, when an active arc is separated from a continent by intra-arc rifting, rifted fault blocks of the original continental forearc will remain beneath the active arc. When west-dipping subduction beneath the Caribbean crust began, apparently by the Aptian according to the time of Caribbean HP metamorphism, the subduction zone, like today, must have been situated east of the Aves Ridge; that is, within the Colombian back-arc basin. Subsequently, the Aves Ridge received Late Cretaceous arc magmatism during eastward migration with the Caribbean Plate (Neill et al., 2011), and prior to the Paleogene opening of the Grenada Basin. Furthermore, the Juan Griego unit on Margarita has a continental protolith of Paleozoic age. If Margarita was separated from the Aves Ridge during opening of the Grenada Basin (Pindell and Kennan, 2009), then the Juan Griego may exemplify other crustal slivers lying below the Aves Ridge. This model (Maresch et al., 2009) makes the implicit prediction that the western margin of the Aves Ridge is a paleo-trench of Late Jurassic-Aptian age. I will present a seismic line across this margin that was recently reprocessed by ION Geophysical, and which appears to show a long-buried accretionary prism where the model predicts it should be.

AAPG Datapages/Search and Discovery Article #90330©2018 AAPG Hedberg: Geology of Middle America – the Gulf of Mexico, Yucatan, Caribbean, Grenada and Tobago Basins and Their Margins “Integration of the geology of the area from the southern onshore of North America to the northern onshore of South America”, Siguenza, Guadalajara, Spain, July 2-5, 2018