ASIA/PACIFIC: All hydrocarbon formation elements present in Southern Gulf of Papua

Location map showing play fairways.

The Gulf of Papua contains a prospective, but lightly explored, rift basin. This basin orig-inated in Jurassic time as the southern portion of the larger Papuan basin rift system, and belongs to the family of productive Mesozoic rift basins that rim the northern margin of the Australian continent.

Seismic data suggests that structuring in the southern Gulf area is mostly extensional and involves basement. We have mapped many potential fault traps with structural relief at reservoir levels that are generally less than 50 meters, but ranges up to 500 meters. Relief generally decreases from the basement upward through the Jurassic-Early Cretaceous reservoirs, and relatively little relief exists above the base Tertiary unconformity. Thus, structural closures were available for hydrocarbon accumulation shortly after deposition of the Jurassic-Lower Cretaceous reservoirs.

Based on nearby well control, multiple high-reservoir-quality Jurassic and Lower Cretaceous sandstones are present. Toro and Imburu sandstones are key targets, as in the thrust belt. Lower Cretaceous Alene sands may be prospective, and in some structural closures, may be associated with seismic amplitude anomalies.

Source rocks in the Gulf of Papua are of comparable quality, but are thicker than source rocks in many of the productive Northwest Shelf basins. Geohistory modeling suggests that petroleum has been generated during two pulses, near end-Cretaceous time and today. The end-Cretaceous pulse probably affected a larger volume of source rock than today's pulse. Because of the predominant east dip of the basin throughout its history, migration paths during both pulses resulted in westward migration paths to the numerous moderate-depth traps located in the reservoir fairway.

As with any rank wildcat area, there are great uncertainties, but our evaluation suggests that significant oil and gas fields may underlie the southern Gulf of Papua. This work results from Arco's exploration program on Petroleum Prospect Licenses (PPLs) 182 and 207 in the southern Gulf of Papua from 1996-1999.

Geologic setting

Seismic section across the southern Gulf of Papua.

The Gulf of Papua basin belongs to the family of productive Mesozoic rift basins that rim the northern margin of the Australian continent. These basins typically contain more gas than oil, because their marine Middle-Upper Jurassic or Lower Cretaceous source rocks have mixed gas-oil potential and are in many areas over-mature. Although the quality of these source rocks is modest, typically averaging 150-300 mg HC/g rock HI and 1-2% TOC, they have nevertheless produced large fields on the Northwest Shelf due to their unusually great thickness.

Worldwide, source-rock thicknesses average less than a few hundred meters. In contrast, our mapping indicates that source-rock thicknesses in much of the Browse basin exceed 1 km, and in the Vulcan graben typically range from 0.5 km to 2.0 km. In some of these basins, lateral migration distances are long, because the carrier beds are laterally persistent and simply structured. For example, gas in Wiriagar Field in the Bintuni basin migrated about 50-100 km.

Well data and a 2D reflection seismic grid with an average spacing of 2 km demonstrate that a thick, north-south oriented Mesozoic basin is present in the Gulf of Papua. This basin may have originated in Triassic through Mid-Jurassic time as the southern portion of a larger rift system that included the area of the present-day thrust belt to the north. The entire margin was passive by Early Cretaceous time.

During the late Cenozoic, the northeast portion of the rift system underwent significant shortening that resulted in the productive Papua New Guinea thrust-belt structures. However, the Gulf of Papua portion of the rift escaped shortening and, as a result, retains the pre-shortening rift-system geometry. Thus, the Gulf of Papua offers exploration opportunities in a proven petroleum system without the structural complexity prevalent in the thrust belt.

The prospective closures in the southern Gulf resulted from rifting or reactivation of rift-related structures. On PPLs 182 and 207, we have mapped numerous basement-involved, probably extensional traps. Consistent with the basin history outlined above, structural relief decreases strongly from basement upward through the Toro section, and the pre-Tertiary unconformity is generally mildly deformed. Thus, structural closures were available for hydrocarbon accumulation shortly after deposition of the Jurassic-Lower Cretaceous reservoirs. In the southern Gulf of Papua, little shortening is observed, unlike the thrust belt only 200 km to the north.

Geochemical and geological data suggest that Middle-Upper Jurassic strata are important sources of oil and gas for many Papuan basin fields. We argue that:

• Middle-Upper Jurassic source rocks in the Gulf of Papua have significant generative capacity.
• The quality of these source rocks is similar to that of Jurassic-Lower Cretaceous source rocks in productive Northwest Shelf basins
• These source rocks have generated substantial volumes of oil and gas.

The Upper-Middle Jurassic throughout most of the Gulf of Papua has fair to good source-rock richness. The best-quality source rocks occur in the northern Gulf and fair-to-good quality source rocks are present in the southern Gulf in fetch areas draining to prospects within PPLs 182 and 207. Yields in most of the Gulf are 2 g to over 4 g HC/kg rock. These are comparable to yields for rocks that have sourced mixed oil and gas for fields on the Northwest Shelf. For example, geochemical data from Sahul syncline source rocks indicate average yield values of 2-5 g HC/kg rock. Average yields in Vulcan graben source rocks are mostly 3-5 g HC/kg rock.

Total source-rock potential in the Gulf of Papua is moderate to high, with a source potential index mainly of 5-15 metric tons HC/sq meter. These values are higher than those of the Sahul syncline (2-5 metric tons HC/sq meter) and the Browse basin, and about the same as values in the Vulcan graben (3-25 metric tons HC/sq meter). The high SPI values in the Gulf of Papua result from the great average thickness (2-3 km) of the Gulf Middle-Upper Jurassic source rock.

Maturity

Our maturity studies have resulted in two broad conclusions:

• That the Jurassic source rocks in the Gulf of Papua have generated petroleum in two discrete pulses, the first at the end of the Cretaceous and the second at the end of the Cenozoic
• That of the two pulses, the end-Cretaceous pulse was volumetrically more important.

Maturity data from the Dibiri-1 well show a significant jump at the base Tertiary unconformity and possibly in some other wells. The magnitude of the maturity jump is greatest where the most Jurassic-Cretaceous section is removed beneath the base Tertiary unconformity, which is why the jump is so obvious at Dibiri-1. The Dibiri-1 geohistory model explains the jump in the maturity profile, and shows both an end-Cretaceous generation event and a late-Cenozoic generation event.

Significantly, in the Dibiri-1 model and in similar models at Anchor Cay-1, Wabuda-1, Kusa-1, and Goaribari-1, most of the petroleum generation occurred at the end of the Cretaceous and not during the late Cenozoic. Typically, the increase in hydrocarbon transformation ratio in the late-Cenozoic generation pulse is less than 5%. Although this late-Cenozoic pulse was much smaller than the end-Cretaceous pulse, it affected a large volume of source rock, so could still result locally in significant petroleum charge.

We infer that oil and gas was generated in the bottom of the source rock over large areas of the Gulf during the end-Cretaceous generation pulse. The middle and upper parts of the source rock were of course less mature but also underwent widespread generation. We base these inferences on our reconstruction of base source-rock paleostructure and oil- and gas-window depths at the end of the Cretaceous.

How was this highly interpretive map made? We used the paleostructural depth points for the base of the Jurassic source rock at end-Cretaceous time from geohistory models and guided our paleostructural contouring using seismically derived thickness maps of the preserved Mesozoic section. Also, preserved Mesozoic thicknesses in certain onshore wells usefully constrain minimum end-Cretaceous depths, so these were also used as data points. The oil- and gas-window depths were based on the geohistory models.

Migration

The eastward dip of the Mesozoic strata in most of the Gulf of Papua favors long-distance migration for any hydrocarbons generated in the deep basin up to the shallow west side of the basin. The Oriomo high plunges eastward into the basin, and may have enhanced focusing of petroleum onto PPLs 182 and 207. We hypothesize stratal migration paths through Koi Iange, Imburu, and Toro Formation sandstones and across and up the many normal faults in the basin.

Our maturity and structure maps predict that the amount of Mesozoic oil draining into Tertiary reef traps in the eastern Gulf is limited, because:

• The reefs were not present when the important end-Cretaceous generation pulse occurred.
• Most of the oil-generation potential was exhausted during this early pulse, leaving in general gas-condensate generation potential for late Cenozoic generation and migration.
• Cenozoic migration paths within the Mesozoic section are mainly westward, away from the reefs. Thus far, large volumes of Mesozoic-sourced oil have not been found in these reef traps.

Numerous high-quality reservoir targets are present in the southern Gulf of Papua. Toro and Imburu Formation sandstones are of similar thickness in the southern Gulf as they are in the thrust belt, where they are the principal reservoirs. In the southern Gulf, as in the thrust belt, these sandstones were deposited in a marine shelf environment. This is evidenced by paleontologic data, common chamosite/glauconite presence, extreme lateral continuity, relatively coarse grain size, and extensive burrowing observed in Kusa-1 conventional core). As suggested by the generally high sand content of the Jurassic and Lower Cretaceous section in Mutare-1 and Iamara-1, the Oriomo high may have been a local source of sand through Jurassic and even up into Early Cretaceous time.

Other formations also contain thick sandstone targets. At PPLs 182 and 207, the possibility of sandstone in the Alene (base Ieru) interval is noteworthy because seismic amplitude anomalies are observed in some of our prospects in that interval. A thick sandstone target in the basal Koi Iange (?) Formation in Magobu Island-1 has long been noted. Reservoir quality in the southern Gulf is excellent. Log-calculated porosities are high. Toro and Imburu Formation sandstones and the nearly 100 meter-thick basal Koi Iange sandstone in Magobu Island-1, Wabuda-1, and Kusa-1 wells have porosities well above 20%. Generally low gamma ray log values (30-50 API units), cuttings descriptions, and a petrographic study of Wabuda-1 indicate that these sandstones are clean.

Dry-hole analysis

Wells drilled to date in the Gulf of Papua have not adequately tested the potential of the Mesozoic. Most failed either because they penetrated no reservoir rock or were outside structural closure. Adequate reservoir sands have not yet been penetrated in the east, which explains the dry holes there.

Why were so many wells drilled outside of closure? There are three possible explanations:

• Wells on the Oriomo high (Wohomul-1, Maremosab-1, Wuroi-1, Mutare-1, and Iamara-1) were drilled with little or no seismic control
• A few, such as Borab-1 (Phillips, 1968), were sited to test overlying Tertiary reef prospects and are unfavorably located to test Mesozoic traps
• Several were drilled on seismic time highs in areas of very low structural dip but where shallow lateral velocity variation is common.

The Morigio-1, Rama-1, Wabuda-1, Anchor Cay-1, and perhaps the Magobu Island-1 dry holes are interpreted in this way. Stacking-velocity analysis suggests that Morigio-1 and Rama-1 are not depth closures, but are time pull-ups beneath high-velocity Tertiary carbonate. Wabuda-1 was drilled on an apparent time high that disappeared after reprocessing by Arco using refraction statics. The apparent high was caused by a velocity anomaly within the Quaternary section, a common problem in the Fly River Delta. Magobu Island-1 is difficult to evaluate because of poor seismic in the Delta.

Prospective traps

Numerous structural traps have been mapped on PPLs 182 and 207. Most traps involve normal faults on one or more sides and four-way dip closure is uncommon. Fault throws are typically 20-150 meters, although at I prospect, the maximum fault throw is 500 meters.

It is believed that fault traps, even if multi-sided, can be commercial in the Gulf of Papua. High shale percentages favor cross-fault sealing for fault traps. Overall, the lower Ieru through Koi Iange interval is mainly shale. Also favoring commerciality is the possibility of multiple stacked pay sands. There are several objective formations, and each formation might have multiple pays sealed intraformationally.

Seismic amplitude anomalies at potential reservoir levels in certain prospects correlate with structural highs. Additionally, Class 2 AVO anomalies were observed on selected reprocessed lines over the M prospect. Could these stacked amplitude and AVO responses directly indicate hydrocarbons? To answer this question we undertook seismic modeling at Wabuda-1.

Ninety kilometers from the prospective area is the nearest well with a modern log suite that penetrates the sands of interest. This well is stratigraphically and structurally on strike with the prospects. A suite of models was generated which tested all reasonable combinations of fluid types and properties (wet, oil with GOR from zero to 1,000 scf/BBL, and dry gas), shale velocity (3,050 - 3800 mps), and sand geometry (sharp or transitional contacts).

Seismic modeling suggests that bright stacked amplitude and positive Class 2 AVO response may be sandstone containing high-GOR oil or dry gas. Dim stacked amplitude or negative AVO response may be wet sand, low-GOR oil sand, gas sand or high-GOR oil sand with transitional contacts, or poor seismic data quality.

In other words, the lack of direct hydrocarbon indicators does not indicate absence of hydrocarbons. The bright amplitudes at M prospect are encouraging and fit all of the seismic criteria to be gas or light oil. However, these conclusions must be considered preliminary because there is little experience with direct hydrocarbon indicators in the Papuan basin.

Diagenetic zones

Hydrocarbon-related diagenetic zones (HRDZs) like those depicted on the Northwest Shelf may be present within Tertiary carbonates on PPLs 182 and 207. On the Northwest Shelf, HRDZs are expressed as vertical zones of poor seismic data quality under which reflections are pulled up. They are high-velocity, irregularly shaped hydrocarbon-leakage-related mineralized zones that overlie faulted oil and gas accumulations. The possible HRDZs on the two PPLs are not randomly distributed, but instead are associated with the most intensely faulted prospects (H, I, N). If these are HRDZs, their presence suggests that petroleum has accumulated in or at least migrated through these prospects. Several exploration risks exist for the southern Gulf of Papua. In order of importance they are:

• Were oil pools that formed during the Cretaceous generation pulse biodegraded due to substantial uplift during the Paleocene? Some of the pools formed at end-Cretaceous time were uplifted to within 200-500 meters of the surface as the base Tertiary unconformity developed. Of course, many examples of light oil fields at shallow depths can be found.
• Did significant amounts of hydrocarbon leak after pools were formed? A number of fields on the Northwest Shelf lost part or all of their hydrocarbon columns in this way.
• Does migration shadowing occur, as is common in systems involving long-distance migration? Detailed structural mapping indicates that certain prospects on PPLs 182 and 207 may be in migration shadows, possibly limiting charge from the northeastern kitchen. If true, charging would be more dependent on the southeastern kitchen.
• Is the source kitchen to the southeast of PPLs 182 and 207 as large and are the source facies there as oil prone as we have mapped?
• Is erosion beneath the base Tertiary unconformity deeper than we have interpreted, removing some of the objective reservoirs, especially the highest ones?

We conclude that the Mesozoic section in the southern Gulf of Papua is prospective. It can be plausibly argued that each element required for oil and gas field formation is present - source, seal, reservoir, trap, appropriate timing, and migration path.

Wells drilled to date in the Gulf of Papua have not adequately tested the potential of the Mesozoic. The Jurassic source rocks in the Gulf of Papua are comparable in quality but are generally thicker than the source rocks in the productive Browse, Vulcan, and Sahul basins on the Northwest Shelf

The Jurassic source rocks have mainly reached oil-window rank. However, we expect that at least as much gas as oil has formed due to the mixed gas-oil character of the source rocks.

Much more oil was generated in the southern Gulf of Papua at the end of the Cretaceous than in the late Cenozoic. Maturity indicators measured today in the Gulf of Papua in many places reflect an end-Cretaceous generation event, and may not indicate that generation is occurring today.

The most important exploration risks are biodegredation during early Tertiary uplift and post-accumulation hydrocarbon leakage.

Reference

4th PNG Petroleum Convention (Published with the permission of the PNG Chamber of Mines and Petroleum and the Department of Petroleum and Energy of the Government of Papua New Guinea. Arco has now surrendered the licence area).