Exploration to first production on Block 9 off South Africa

SOUTH AFRICA OFFSHORE E&P-1

Paul L.A. Burden, Christopher P.N. Davies
Soekor E&P (Pty.) Ltd.
Cape Town

Offshore Block 9 has been the focus of South African hydrocarbon exploration for the past 17 years.

This 23,000 sq km block covers most of the Bredasdorp basin, a sub-basin of the Outeniqua basin, which extends over most of the southern offshore (Fig. 1) [54,334 bytes]. Block 9, which includes Oribi oil field, is operated by Soekor E&P (Pty.) Ltd. Joint ventures are currently available in several areas within the block.

Sediments in the Bredasdorp basin range in age from Jurassic to recent (Fig. 2) [30,512 bytes].A late Jurassic to early Cretaceous, fluvial to shallow marine rift fill is overlain by a marine drift succession deposited after Gondwana breakup. The latter includes substantial thicknesses of deep marine sediments that have been separated into second and third order sequences using seismic/sequence stratigraphic techniques.

One of these, the 14A sequence, is a third order sequence of Albian age (Fig. 2), sourced from the west of the basin. Its lowstand tract, developed in the central Bredasdorp basin, consists largely of a submarine fan complex that includes numerous channel and fanlobe sandstones and overbank fines encased in deep-marine shales and has proved to be the most prolific oil play in the basin to date.

The area of the play lies between 80-120 km offshore in 100-120 m of water in the Indian Ocean. The target reservoirs are encountered between 2,350 m below mean sea level in the west and 2,750 m below mean sea level in the eastern part of the play area.

Exploration history

Exploration of the 14A lowstand tract began in 1987 after an oil discovery in the western part of the play area at the E-AA1 borehole (Fig. 1).

Deep-marine massflow sandstones 44 m thick were intersected at the 14A level, the top 10 m of which were saturated with light oil that flowed at over 4,000 b/d on test with a gas-oil ratio (GOR) of 580 scf/bbl. Similar oils were discovered at the E-AD1 and E-AR1 boreholes during 1987-88. E-AR1 flowed about 5,000 b/d for 30 days with no sign of depletion. This test was the first convincing evidence that sustainable oil production with good pressure support from a large aquifer was possible from these sandstones.

In 1990, the E-BT1 borehole intersected 32 m of sandstone, 27 m of which were oil saturated. Average porosities of 18% and permeabilities of up to 2 darcies were measured. The borehole flowed 8,730 b/d of oil. This accumulation, renamed Oribi oil field, is now in production and will be discussed in a second part next week.

In the eastern part of the play area, gas and oil were discovered in E-AJ1 in 1988 followed by a gas discovery at E-BK1 and gas and oil at E-CB1 and 2 and E-CR1.

To date, 46 boreholes have intersected this play (Fig. 1). Sandstone reservoirs were intersected in 28 (61%) of these and have been classified as follows:

Hydrocarbon saturation 54%
Shows 18%
No shows 29%.

Presently, exploration activity is centered around Oribi field, within the reach of the Orca floating production facility, and in the southeastern part of the play around E-CB gas/oil field, where a number of prospects have been mapped.

It is intended that the Orca facility will eventually be relocated and used for production of other economically viable 14A fields as water depths and sea conditions are similar throughout this play area.

Data coverage

Seismic data

More than 45,000 km of good quality, multichannel seismic 2D data are available over the Bredasdorp basin. Coverage is dense over the whole of the play area. Many of the older vintages of data were recently reprocessed, improving their resolution significantly.

In addition, about 1,300 sq km of 3D seismic data are available in two large data volumes covering most of the western and southern parts of the play area (Fig. 3)[45,017 bytes].

Borehole data

All boreholes in the area have full suites of geophysical logs and full sets of sediment samples.

High resolution sequence stratigraphic interpretation using planktonic and benthonic foraminifera species abundance and diversity curves has been carried out at the 14A level for most of the boreholes, and this has been used for detailed correlation at subseismic scale.

Thousands of Rock-Eval, Leco pyrolysis analyses, and optical analyses including kerogen maceral, fluorescence, and vitrinite reflectance studies have been done, and several hundred oil and source rock bitumen samples have been analyzed by extraction and gas chromatography. More than 100 of these have been subjected to gas chromatography-mass spectrometry analysis of biomarkers.

Petrographic studies of thin sections including SEM work have been carried out on all of the porous 14A sandstone intervals. These data have been used in detailed borehole correlations as well as porosity and permeability distribution studies.

Other data

A regional Sniffer survey of all offshore areas was acquired in 1986, about 2,430 line km of which are in the Bredasdorp basin.

This survey highlighted areas around the basin margin where hydrocarbon seeps occur indicating that migration is still taking place in the basin today.

14A seismic expression

The stratigraphic relationships of the 14A sequence are illustrated on seismic lines running east-west (Fig. 4) [29,710 bytes].

The 14A lowstand tract is distinguished by a series of high amplitude reflections at the base of the 14A sequence. This tract laps out just updip of the base of slope on the basal type1 unconformity, 14At1. It is overlain by a thicker progradational highstand tract. The sequence thins toward the east.

Most reservoirs in the 14A lowstand have low acoustic impedance seismic responses caused primarily by their low densities and high porosities. On standard negative polarity (Fig. 5) [55,091 bytes] the top of the sandstone corresponds to a peak and the base to a trough.

Seismic amplitudes are often higher where hydrocarbons are reservoired, especially in the case of gas where bright spots may be present, for example over the E-BK gas reservoir (Fig. 6) [50,988 bytes].

Soekor has developed an AVO technique known as Geostack. Distinct Geostack anomalies are associated with all discoveries with hydrocarbon columns greater than 5 m, including Oribi oil field and E-CB oil-wet gas field (Fig. 7) [23,845 bytes].

Geostack has thus proved to be a reliable direct hydrocarbon indicator. The amplitudes of the anomalies are roughly proportional to the GOR of the reservoired hydrocarbons, the higher the ratio the stronger the anomaly, and as the oils in 14A reservoirs are generally light with GORs greater than 500 scf/bbl the technique works well in these reservoirs.

Bredasdorp geology

The source for the sediments of the 14A sequence is in the western area of the Bredasdorp basin (Figs. 4 and 8).

During the early Albian, a relative sea level fall led to the formation of the basal unconformity, 14At1, which eroded into the pre-existing highstand shelf sandstones in the western basin resulting in the deposition of the sand-prone 14A lowstand tract in the center. The lowstand tract is up to 250 m thick in the west, while towards the east beyond the E-AJ prospect it thins to less than 50 m.

Turbidity currents transported this material into the basin from the west, depositing sandstones and siltstones within numerous submarine channels, some of which terminated in extensive fanlobes (Fig. 8) [32,027 bytes].

The fanlobes have a coarsening upward or a blocky gamma ray log motif, while channelized reservoirs are blocky or fining upward. Some lobes are capped by channel abandonment facies (Fig. 9) [24,938 bytes]. Fanlobes generally have more shale interbeds. Channelized sandstones are more numerous in the western part of the play area, while in the eastern area fanlobes predominated (Fig. 8).

The sandstones are generally very fine to medium grained, clean to slightly argillaceous and glauconitic. Reservoir intervals are made up of numerous stacked massflow units and vary in thickness from as little as 5 m to over 100 m. Most of the hydrocarbon discoveries are in reservoir successions of 30 m thickness or more. Areas of thick sandstone deposition are often fringed by areas of stacked low energy, low density, distal turbidite deposition.

The play is not directly affected by large scale faulting. However, faulting on a scale close to the seismic resolution (throws of less than 20 m) controls reservoir distribution and continuity in places.

The majority of the hydrocarbon accumulations in these reservoirs appears to be trapped as a result of a combination of complex compactional, tectonic, and sedimentary mechanisms.

Traps that are characterized by four-way domal closure on the top of the sandstone but not on the base are thought to have formed as a result of differential compaction over sandstone depocenters. Other traps characterized by domal closures on both top and base are thought to have been formed by either late structural inversion or compactional drape over prominent structures beneath them.

Some of the hydrocarbon accumulations occur within easterly plunging noses with no apparent structural closure on their updip (western or southwestern) sides. In some cases the formation of the noses can be attributed to drape over deeper structure, but the updip trapping is attributed to updip pinchout or abutment against pre-existing fault scarps.

Evidence from pressure and salinity trends suggests that many of the individual sandstone bodies are in hydraulic connection due to the high sandstone content of the turbidite system. It is therefore likely that reservoirs will enjoy pressure support from a large common aquifer during production, as indicated by the E-AR1 extended well test.

A regional study of formation water salinities shows that the reservoir trend was flushed by meteoric water, probably during the late Tertiary, and subsequently partly recharged by connate water (Fig. 10) [39,735 bytes].

Geochemical aspects

Source rock distribution

The most widespread known oil prone source shales in the central Bredasdorp basin lie between 100 m and 300 m beneath the 14A lowstand reservoirs in the deep marine, transgressive, and early highstand tracts of the Aptian, 13A sequence (Fig. 2). These shales have an average thickness of more than 90 m over an area of about 3,300 sq km (Fig. 11) [19,097 bytes].

Total organic carbon (TOC) averages 2.8% and can exceed 4% locally. Visually, almost all of the kerogen is Type 2, amorphous, fluorescent kerogen with hydrogen indices (HI) in the order of 450-600 and the capability of expelling 10-15 kg of oil/tonne of rock. Extract and pyrolysis evidence for oil expulsion is common, and numerous shows of reservoired oil above and below the shales confirm the potential.

A number of thin intervals of gas-prone source shales have been intersected in the Barremian to Hauterivian succession (Fig. 2). These have TOC contents of over 2.5% and HI values of about 300. They have the potential to have expelled 6-8 kg wet gas and condensate/tonne of rock.

Thermal maturity data from each borehole (vitrinite reflectance and Rock Eval pyrolysis) indicate that the 13A source shales are almost wholly in the oil window, while the Barremian-Hauterivian source rocks are presently in the gas window. Present day geothermal gradients vary from 3.8° C./100 m in the west to 4.4° C./100 m in the east.

HC types, migration

Three different hydrocarbon types have been encountered in 14A reservoirs:

Low maturity, heavy residual oils present in traces in boreholes in the southeastern play area where wet gas is presently reservoired.
Chemically similar but higher maturity light oils (35-45° gravity) in the western (updip) portion of the play area, including that reservoired at Oribi and other fields in the area.
High maturity wet gas and condensate (45-55° gravity) in the southeastern play area including E-BK gas field and E-CB gas cap.

Biomarker data from reservoired oils and oil traces match biomarker data from the 13A (Aptian) source shale from up to a few kilometers down dip, suggesting that migration took place over fairly short distances. Faults and fractures are thought to have acted as conduits for the migration.

Burial history modeling shows that two phases of expulsion from these source rocks occurred at about 60 Ma and 5 Ma. The lower maturity, residual oils are interpreted to have been expelled during the first expulsion phase and the higher maturity oils during the second.

The maturity differences between the two types can be accounted for by the difference in the burial depth of the 13A source rocks at the time of the first expulsion as compared to the second.

Biomarker and isotopic studies indicate that the wet gas and condensate are from a different source rock that contains large amounts of terrigenous kerogen. The distribution and depth of burial of this source is unclear, but it could be within the Barremian-Hauterivian early drift sediments. Migration of gas from this source into 14A reservoirs must have taken place after the first phase of oil migration as the gas has clearly displaced oils from some of the reservoirs in the southeastern parts of the 14A play area, for example, E-BK gas field.

Next week: Exploration, appraisal, and start-up of Oribi oil field