Editor’s Note: This is the second article in a three-part series on the history of deepwater exploration. The first article, Deeper Waters: How Science and Technology Pushed Exploration to Greater Depths, from the January 2017 issue of Explorer, explained the fundamental paradigm shift that deepwater reservoirs exist. The third article, covering above-ground issues with deepwater E&P, is planned for 2021.
During the 1980s, early prospecting in deepwater margins was the simple extension of prolific updip producing basins that were charged (Gulf of Mexico and Campos basins, then later Nigeria, Angola/Congo, Baram and Mahakam, see figure 1). However, the migration of petroleum exploration and development into deepwater (>1,500-feet water depth) was fraught with challenges. For engineers and geologists, one of our biggest challenges was to overcome our collective lack of understanding about the geology of deepwater. We had to revisit everything we thought we knew, one paradigm shift after another.
Our first paradigm shift, as mentioned in Part 1 of this series, was to realize that petroleum reservoirs actually existed in deepwater. According to conventional wisdom at the time, “There’s just mud at the bottom of the Gulf of Mexico … nothing’s going on out there.” But with new, albeit low-resolution seismic sampling technology, we learned that reservoir sands actually were deposited from sediment gravity flows in deepwater settings. Furthermore, we suspected that these deepwater reservoirs might be wildly profitable, under the right economic conditions.
The second paradigm shift was working out the details of the evolution of continental margins (plate tectonics), and where petroleum systems fit into this framework. How did the elements of land-based and shallow-water-based petroleum systems – source rocks, top seals, traps and generation/migration/accumulation – translate to the murky ocean depths? Would we be able to find any source rocks? Did adequate top seals even exist underwater? The timeline in figure 2 summarizes the petroleum systems elements that had to be understood before companies would be willing to risk deepwater exploration and production.
The November 2016 Historical Highlights article in the EXPLORER reviewed the historical development of plate tectonics as a discipline. Understanding the geologic evolution of the continental margins was critical to unlocking the secrets of the petroleum systems in deepwater margins. In fact, exploration in deepwater has significantly contributed to the geoscience community’s growing understanding of the evolution of continental margins.
Five new concepts were essential for industry’s move into deepwater:
- There are three margin types – divergent, convergent, transform – and each has a unique distribution of petroleum systems elements and hydrocarbon potential.
- Divergent margins evolve from initial rift basins into mature, fully developed ocean basins with active oceanic spreading centers.
- Different crust types are distributed differently, with regional variations in heat flow.
- Basement tectonics has controlled the location of lacustrine source rocks, some overlying depositional systems (e.g. early carbonate buildups), and some of overlying structural features and their location.
- The stratigraphic evolution of continental margins is predictable through time, in terms of both sediment type and caliber. Understanding the stratigraphy can help predict the distribution of petroleum systems elements, including source rocks, seals and reservoirs.
Curiously, in the 1960s, several exploration wells were drilled along the shallow-water portion of convergent margins, in either forearc basins or subduction-related rocks of offshore Oregon, Washington, British Columbia, Mexico and Nicaragua. The wells targeted structures with four-way closures as defined on 2-D seismic. The wells were drilled before the three kinds of plate boundaries were recognized. Thus, these wells were drilled with incomplete knowledge of reservoir presence and quality, source rocks, heat flow and structure. Since that time, exploration activity has changed its focus to primarily divergent and transform margins, and in a few cases, contractional margins (Colombia, northwest Borneo).
Initially, some companies doubted that oil-prone source rocks were present in deepwater margins. This belief was due to incomplete data about the evolution of early continental margins, the oceanographic conditions responsible for source rock accumulation, and the temporal and spatial distribution of source rocks today.
For several millennia, we already knew that oceans had oil seeps. For example, written records BCE indicate oil accumulating on the hulls of Egyptian and Roman boats. Seeps were also used to waterproof hulls – the hull of Noah’s ark being one famous example (Genesis 6:14). During the early European exploration of the western hemisphere in the 16th century, the diaries of Hernando DeSoto and Francis Drake report oil on the hulls of their ships in the Gulf of Mexico and coastal California, respectively.
But where were the oil seeps coming from?
Four observations, starting in the 1960s, helped to answer that question (figure 2):
• The cores from the Deep Sea Drilling Program (and later, to a lesser degree, Ocean Drilling Program, Integrated Ocean Drilling Program) conclusively recovered source rocks and sealing lithologies. These findings became essential for successful exploration in deepwater.
Specifically, the early drilling sites of the DSDP (1968-83) recovered cores in many potential source rock facies with high total organic carbon and hydrogen index. For example, the cores from Site 2 of DSDP Leg I (central deep Gulf of Mexico), demonstrated shales with high TOC values and oil in the cores (which caused the crew to abandon the hole quickly).
Today, of the more than 2,000 drilling sites, about 200 have penetrated black shales with relatively high TOC (according to Steve Bell in a personal communication in 2016). The presence of these potential source rocks in deepwater was a critical factor in reducing the exploration risk of deepwater plays.
• In 1976, Seymour Schlager and Hugh Jenkins recognized the development of global ocean anoxic events in Cretaceous and Jurassic strata, largely based on the integration of results from multiple DSDP sites. These are critical intervals for source rock deposition in many deepwater basins globally. In addition, several more local anoxic events were recognized in the Lower Cretaceous and possibly Jurassic strata. When these findings were correlated to well-known outcrops, geoscientists obtained a more unifying temporal and spatial understanding of source rock deposition. With this knowledge, we could better predict the possible presence of source rocks in basins.
• In 1941, C. H. Pan was the first to recognize that lakes can contain prolific source rocks, specifically in the northern Shaanxi Province of China. Independently, many workers realized that lakes commonly develop in rift basins, both modern and ancient. By the early 1970s, with improved seismic imaging and plate tectonic frameworks, rift basins were recognized to rest in ultra-deepwater along some margins.
The first deepwater exploration area where lacustrine strata were demonstrated as the primary source rock for oil and gas fields was in the southern Atlantic basins. The significance of these strata as potential source rocks was recognized by a group of Petrobras geoscientists (Egon Meister, Lincoln Guardado and others), while working on early offshore discoveries in the Campos Basin in the late 1970s and ‘80s (such as Namorado, Enchova, Pampo, Garopa). Based on oil types and rock samples, these geologists were able to recognize the Barremian (Lower Cretaceous) Lagoa Feia as the source rock. Later, in the mid-1980s, the oils recovered in the Petrobras Marlim and Albacora deepwater fields confirmed the presence of lacustrine source rocks.
Petrobras did have one advantage in their early work – the first field discovered onshore 1954 in the Bahia Basin was in lacustrine turbidite reservoirs. So, Petrobras had an early indication that lacustrine source rocks could be present in their deepwater margins.
• By the late 1980s, Total geoscientists recognized that marsh plants were the source of oils in their Mahakam Delta and offshore shelf fields along southeast Borneo. Ten years later, plant leaves were found interbedded with reservoir sands in the deepwater discoveries by Unocal. With more geochemical work, the leaves of the Nipa Palm were determined to be the source of the oil along the island of Borneo in deepwater. The high wax content of these leaves preserves the lipids for oil generation (figure 3).
As we discussed in the January 2017 column, it took several decades for geoscientists to recognize the presence and significance of sediment gravity flow sands in deepwater settings. This is mirrored in the rate of new exploration success. Later, additional shallow-marine to continental reservoirs, which were deposited during the early stages of the margin’s evolution, were discovered (figure 1).
Top Seals: Presence and Integrity
Early on, the industry assumed that deepwater contained the fine-grained sediments that create seals. This was confirmed initially by oceanographic institutes that collected shallow penetration piston cores in the 1950s and ‘60s in deepwater settings. Later, the DSDP recovered extensive clay-sized sediments in many cores. Eventually, shallow-marine fine-grained shales (Brecknock, 1979) and carbonates (Malampaya, 1993), as well as lacustrine shales (Tupi, 2007), were found to seal shallow-water and continental reservoirs discovered in deepwater.
In many onshore fields, salt had already been recognized as an effective seal in petroleum systems. Later, in deepwater, salt was first recognized as the updip seal in many fields in the northern GOM. Then, beginning in 2007, most of the fields in the pre-salt play in the Santos, Campos, and Espírito Santo basins were found to have evaporite as the top seal. More recently, evaporite top seals were found in the pre-salt discoveries of Angola and Egypt (Zohr).
As for seal integrity, an ongoing concern with seals is the possibility of failure after initial petroleum entrapment, or in the parlance of industry, “blown seals.” Many exploration wells drilled attractive prospects, only to find shows of petroleumwith or without good reservoirs.These exploration failures generated new approaches to pore pressure analysis. Although the capillary properties of shales did not seem to pose a problem for seal integrity, overpressured reservoirs were found to cause trap leakage via fault planes and/or seal rupture. These failures were often indicated by shallow gas in the overlying reflections (such as chimneys and clouds).
In the 1980s, Shell Oil began applying pore pressure analyses on faulted prospects to assess whether their traps could hold a hydrocarbon column. Early work by Dan Worrall and others linked pore pressure analysis to gas and oil column heights in the Gulf of Mexico. By the early 2000s, pore pressure was being used to define “pressure-protected traps” as well as oil columns where gas had leaked off (for example, Big Bend Field, according to Stephens, 2016).
Acquisition of regional seismic 2-D along continental margins globally began in earnest in the late 1960s and increased during the next two decades. The initial 2-D seismic data imaged many different trapping styles in deepwater, with many possible four-way and three-way closures. Geoscientists were already familiar with these styles of traps onshore; so, their existence in deepwater was easy to accept. However, the seismic data also showed some structural features unique to deepwater: extensive allochthonous salt, mobile shale, and foldbelts (figure 4). None of these features are common in onshore or shelf regions. They were a new paradigm.
Salt was known to provide a trap in many onshore fields and some basins’ shelf. However, the enormous volume and areal extent of salt – much of it allochthonous – along many deepwater margins was one of the early surprises from regional 2-D seismic data. In the shelf of the northern GOM and its upper to mid-slope region, many fields were found along the flanks of salt in three-way closures, faulted three-way closures, and a few stratigraphic traps. Later, at least 24 fields were discovered below the allochthonous salt that have both sub-salt traps and seals. In other basins, 14 fields/discoveries were found in the pre-salt play of Brazil and Angola are sub-salt (figure 1).
Initial exploration efforts in many deepwater basins targeted traps associated with shale features, based primarily on 2-D seismic data. However, with the acquisition of newer, high-resolution 3-D seismic data, we can now re-interpret many features that we thought were shale-related. Specifically, many shale-related features are now recognized to be tightly folded anticlines or thrusted folds without any evidence of shale tectonics. Several fields have been discovered in basins with mobile shale substrates.
Some geoscientists were surprised by the large extent of contractional foldbelts along the base-of-slope in many basins. These foldbelts can have large structural closures.
In divergent margins, foldbelts develop due to overall sediment loading and downslope translation of stress. Initial exploration in foldbelts began in 1987 in the northern GOM, and the first successful discovery was in 1996. Today, fields produce from foldbelt structures in at least five deepwater basins globally.
In collision and transpressional margins, deepwater foldbelts can also develop, often in accretionary wedges. However, these are non-prospective due to a lack of source rocks and/or adequate reservoirs. Foldbelt discoveries have been made in northern Borneo (Brunei and Sabah), Colombia, and drilled in other countries.
One of the big surprises in deepwater exploration was the recognition of the important role of stratigraphic traps. Estimates are that two-thirds of the fields have combination structure-stratigraphic traps, and 10 percent have pure stratigraphic traps. This recognition was important for implementing the best development strategies and trap risk assessment in these discoveries.
Generation, Migration, and Accumulation: Petroleum Systems Modeling
Petroleum systems modeling developed in the mid-to-late 1970s and became routine by the early ‘80s. Later, the fundamental concepts of petroleum systems were formally defined with the publication of AAPG Memoir 60 in 1994. As exploration moved progressively offshore, deepwater evaluations included petroleum systems modeling to reduce risk.
Many factors must be considered when evaluating the generation and migration of petroleum. However, two of these factors dramatically influence the deepwater margins. First, the timing of generation is affected by the kind of underlying crust. Specifically, oceanic crust in deeper-water portions of continental margins has lower heat flow values, so that petroleum generation is delayed.
Second, the presence of extensive salt, both autochthonous and allochthonous, dramatically affects the timing of generation due to differences in heat flow. Given salt’s convective nature, source rocks overlying thick salt tend to generate petroleum more quickly than source rocks that underlie thick salt.
In some deepwater basins with extensive vertical migration, petroleum geochemistry studies indicate multiple stages of entrapment and remigration. These stages were then incorporated into petroleum system modeling.
To restate our summary of January 2017 – what’s next for deepwater exploration and development? As of November 2020, the continued price downturn has had a significant global impact on the economics of deepwater exploration and production. Specifically, many deepwater plays are not profitable in the current low-cost environment due to high capital and operating costs. Clearly an increase in oil price and/or a decrease in operating expenses are essential for deepwater exploration and production to become profitable again in the future.
Capturing the drilling history and facts required input from many colleagues. We are especially grateful to the following:
Sam Algar, John Armentrout, Marco Arreguin, Bert Bally, Sam Bateman, Steve Bell, Joe Besser, Kevin Biddle, Kevin Bohacs, Peter Bond, Chris Cadeau, Eric Childs, Tim Collett, Steve Cossey, John Dolson, Marlan Downey, Flavio Feija, Joan Flinch, Brian Frost, Bob Fryklund, Steve Graham, Lincoln Guardado, Paulo Guimarães, Will Gutterman, Michelle Judson, Hans Krause, Laurie Lamar, Ray Levey, Paul Lillis, Trey Meckel, Bob Mitchum, Webster Mohriak, Chris Morley, John Pepys, Malcolm Ross, Art Saller, Scott Scuffleton, Sandeep Sharma, Mick Shrimpton, Juan Soto, Pete Stark, Gary Steffens, Gabor Tari and Peter Vail.