Seychelles thermal history, hydrocarbon generation traced

Douglas W. Waples
Exploration Consultant
Evergreen, Colo.

Kerry Hegarty
Geotrack International Pty. Ltd.
Brunswick West, Victoria, Australia

Although little exploration has been carried out on the Seychelles microcontinent, there are a number of reasons to believe the area has considerable potential to produce hydrocarbons.

The source rocks responsible for the huge accumulations of tar in Madagascar may well be present in Seychelles.^{1 2} In addition, it is possible that major Mesozoic facies that sourced vast amounts of oil in the Middle East extend into Seychelles, as recent analyses and review have identified oil prone Type II kerogens in shales of early Middle, and Late Jurassic age.^{1 3}

Third, the source rocks for some of the oils in the Bombay High area (India) may also have been deposited along the northeast Seychelles margin during the initial stages of its Early Tertiary separation from India.^3-7

Finally, the existence on Seychelles beaches of abundant tar balls derived from at least three distinct source rocks gives direct evidence of local hydrocarbon generation.⁸

Tectonic history

The Seychelles microcontinent is a moderately large chunk of Proterozoic continental crust that rifted from Gondwana in stages. ^9-11

From Late Permian (225 Ma) until Aptian time (115 Ma), West Gondwana (Africa) gradually separated (at 160 Ma) from East Gondwana (Seychelles, Madagascar, India, Antarctica, and Australia), creating the Somali basin (Fig. lA [175,457 bytes], Fig. 2 [205,817 bytes]). The narrow western margin of the present Seychelles microcontinent was formed during this event.

The Late Jurassic-Early Cretaceous separation of Australia/Antarctica from India was of little significance for Seychelles, but during the Cenomanian-Santonian (95-84 Ma) Madagascar rifted from India-Seychelles along the Seychelles margin. During the Campanian-Maastrichtian (84-65 Ma) Seychelles-India rotated counterclockwise with respect to Madagascar (Fig. lB), forming the Mascarene basin, whose northern margin defines the southern margin of the present Seychelles microcontinent (Fig. 2).

Finally, in the Late Campanian (76 Ma) India began to separate from Seychelles, leaving the microcontinent alone in the Indian Ocean (Fig. 1C). The northern and northeastern margins of the present Seychelles microcontinent (Fig. 2) were formed during this final rifting event, which probably persisted until the Paleocene.

Each of these rifting events was, of course, associated with a thermal event, but the character of each thermal event is distinct. The development of the Somali basin appears to have been normal passive-margin rifting. The rotation of India-Seychelles to create the Mascarene basin, in contrast, occurred along transform faults.¹² The separation of India from Seychelles seems to have been related to Deccan extrusive volcanism, perhaps the result of hotspot activity,^11-13 which peaked between 69 and 65 Ma.¹⁴

Thermal-history reconstruction

In many maturity-modeling studies the ultimate objective is to calculate timing of hydrocarbon generation, types of products (e.g., oil or gas) generated, and the volumes of hydrocarbons generated and expelled. For such calculations to be of value, however, the thermal history used in the modeling must be substantially correct. Reconstructing thermal histories is often difficult, especially in areas like Seychelles that have suffered complex histories and for which few measured control data are available.

In spite of these difficulties we have reconstructed the thermal history in several known and proposed basins within and adjacent to the Seychelles microcontinent.¹⁵ We present some of our results here. In carrying out those reconstructions we have utilized general conceptual models (e.g., heat-flow trends, values, and histories for rift basins in general), local conceptual models (e.g., the proposed geologic evolution of each of the Seychelles basins), and measured thermal-indicator data (vitrinite reflectance [Ro] and apatite fission-track measurements AFTA©). The use of thermal-indicator data is essential to developing constrained thermal-history models, which are much more reliable and instructive than purely conceptual models.

Thermal-indicator data help us reconstruct a variety of thermal, tectonic, and erosional events that can influence thermal histories. Events that we have included in our models for one or more parts of the Seychelles microcontinent include changes in heat flow related to crustal extension and hot-spot development, decreases in temperature due to uplift and erosion, and increases and subsequent decreases in temperature due to hydrothermal activity. The hydrothermal activity was of particular interest in this study, because in some of the areas modeled it turned out to be a major but previously unsuspected factor controlling total thermal history.

Constrained thermal-history models for this study were initially based on Ro profiles from the four exploration wells drilled to date in Seychelles (locations in Fig. 2), and AFTA data obtained on 12 samples from wells and three samples from outcropping basement granites. These constraints were then extended across the region to predict likely thermal and burial histories at ten undrilled locations (pseudo-wells) within the sedimentary basins of the Seychelles microcontinent (Fig. 2).

The basis and application of Ro and AFTA technologies, which are now routine tools in the oil and gas industry, are not discussed here. Readers interested in further information are referred to the published literature.^16-20

Conclusions on timing

In this study, three major thermal episodes have been identified using AFTA. Timing of each thermal episode has been constrained by the overlap in the timing estimates for each individual sample that experienced that event. Each of these thermal episodes is correlated with a major tectonic event in the region, as described below:

• A Mesozoic episode that we observed only in the granite outcrop samples. AFTA results indicate that cooling commenced some time between 180 and 100 Ma. This "Marion" event may have been linked to the passage of the Marion hotspot across northern Madagascar about 130-125 Ma,21 somewhat before the Mascarene basin began to open (Fig. lB).
• A Late Cretaceous-Early Tertiary episode, the "Deccan" event, which is seen in well samples. It is interpreted as the result of extensive Deccan volcanism and probably involved extensive hydrothermal activity.
• The "Mahe" event, a Late Tertiary episode that was observed in results from both well and outcrop samples. It is apparently due chiefly to movement of hot fluids. However, the immature topography of Mahe Island suggests that some of the observed cooling effects may also be attributed to uplift and erosion.

Two older thermal episodes are included in our models, although they have not been recorded in our AFTA data. These events are predicted from our understanding of the probable geologic evolution of the Seychelles microcontinent. The absence of AFTA evidence for these events does not necessarily mean the events did not take place but rather that any traces of those early events have been eradicated, overprinted by subsequent thermal events.

• A Late Paleozoic event (300-260 Ma) related to initial failed rifting within Gondwana.

• A Permo-Jurassic event (250-160 Ma) related to the initial separation of East and West Gondwana and the formation of the Somali basin.

Late Cretaceous thermal events associated with the opening of the Mascarene basin and volcanism along the Amirante Ridge are not recorded in the AFTA data. These events were apparently not important at the sampled localities but may have been significant in other areas.

These various proposed thermal events are summarized in Table 1 [93,543 bytes].

Implications for prospectivity

Maximum paleotemperatures and timing constraints derived from individual AFTA samples, together with the estimates of maximum paleotemperatures derived from the Ro data, were used in concert to constrain paleoheat flow, convective heat transfer in the past, and amount of removed section for each identified event.

Heating near the K-T boundary is believed to be related to the passage of hot fluids through the shallow section at the time of the Deccan event. Hydrothermal activity of this type and magnitude is not surprising in association with such a massive volcanic event. The paleotemperature profile for the Late Tertiary event is also characteristic of shallow transient fluid effects.

Several source rocks, varying in age from Paleocene to mid-Triassic, have been identified in the regions,^{3 10} but not all those rocks are present in all areas. Assessment of the hydrocarbon potential across the region is therefore complex. Nevertheless, the new constraints on paleotemperature provided by both AFTA and Ro data indicate that heating during any of the abovementioned events was sufficient, at least locally, for hydrocarbon generation (Fig. 3A-D [205,817 bytes]). The implications of our reconstructed thermal histories for prospectivity are summarized below for each major basin in the study area.

Somali rift basin play

Peak maturities and high generation rates were attained during the Marion event (Early Cretaceous) for the deeper source rocks and during the Deccan event (K-T boundary) for the shallower ones (Fig. 3A). Locally generation from the oldest source rocks may have even preceded the Marion event as a consequence of normal burial (Fig. 3B).

Minor well shows, UV fluorescence anomalies, and seismic DHIs (direct hydrocarbon indicators) suggest that generation has in fact occurred. Main phases of deformation and formation of structural traps occurred prior to each of these thermal events.

NW Indian Ocean rift basin play

Peak maturities and high generation rates were achieved during the Deccan event. Local evidence for the presence of hydrocarbons includes gas sniffer anomalies (ethane/iso-butane in the southeast, and propane/normal butane/total HC in the northwest) and seismic DHIs. The main phase of structuring is at least as old as the generation event.

Mascarene rift basin play

Peak maturities with high generation rates occurred during the Deccan event, although the oldest rocks generated hydrocarbons during the Marion event (Fig. 3C). Gas sniffer anomalies (ethane/iso-butane), geochemically unique beach-stranded tar on Coetivy Island, ⁸ UV fluorescence anomalies, and seismic DHIs all support the idea of hydrocarbon generation. Both structural and stratigraphic traps predate the generation events.

Late Paleozoic rift basins play

Maturation in the proposed but as-yet unverified source rocks occurred in several steps (Fig. 3D). The most important of these was the Marion event, but additional generation also occurred during the Deccan event and even the Mahe event. The main phase of structuring predates the earliest generation event.

Providence/Farquhar area

Peak maturities and high generation rates were attained primarily through deep burial during the Campanian to Paleogene (~85 to 40 Ma), long after development of the principal structures in the area. Gas sniffer anomalies (propane/normal butane) provide local evidence for the presence of hydrocarbons.

Conclusions

The thermal history of the Seychelles microcontinent has been complex, but many of the details have been unraveled by modeling of thermal histories constrained with vitrinite and apatite fission track data. Several major heating episodes have been responsible for maturation of a variety of proposed and proven source rocks at different times. In most cases maturation was achieved by conductive heating during periods of high basal heat flow, but convective heating associated with hydrothermal activity has also been locally very important.

Three significant heating pulses have been identified as having been responsible for source rock maturity in the Seychelles microcontinent, of which two previously either were not considered to be significant (the Marion event) or were not recognized at all (the Mahe event). The latter Mahe event is significant in that it induces at least marginal maturity in the Paleocene shale, one of the richest potential source rock horizons identified in the wells. This Paleocene source is equivalent to one of the source rocks for oil in the Bombay High oil province off the west coast of India.^4-7

Results of this study significantly upgrade the oil potential of the Seychelles microcontinent. We have shown that the source rocks have experienced higher temperatures than is commonly believed, leading to higher levels of maturity and hydrocarbon generation. Hydrocarbon generation appears to have occurred in all or most of the basins on the margin of the microcontinent. Much of the generation occurred during the Cretaceous or Early Tertiary, after traps had formed. Preservation of much of the oil we believe has been generated should not have been overly diffcult.

Additional information on the thermal history of the Seychelles microcontinent is available via the Internet at www.geotrack.com.au

References

1. du Toit, S.R., Kidston, A.G., and Slind, O.L., Hydrocarbon potential of the East Africa continental margin, Alconsult International Ltd. report, 1997, 132+ pp.
2. Kamen-Kaye, M., Mozambique-Madagascar geosyncline-I. Deposition and architecture, J. Petrol. Geol., Vol. 5, 1982, pp. 3-30.
3. Plummer, P.S., Joseph, P.R., and Samson, P.J., Depositional environments and oil potential of Jurassic/Cretaceous source rocks within the Seychelles microcontinent, Mar. Petrol. Geol., Vol. 15, 1998, pp. 385-401.
4. Banerjie, V., Mathur, P., and Mathur, M., Geochemical exploration efforts in Bombay offshore basin, in Integrated Exploration Research: Achievements and Perspectives, conference proceedings, J. Pandey and V. Banerjie, eds., KDMIPE India, 1991, pp. 387-393.
5. Pande, A., Khan, M.S.R., Singh, B.P., Haraza, P.N., Nain, K., Tripathi, G.K., Misra, K.N., and Chandra, K., Geochemical index fossils in crude oils in Indian sedimentary basins, Indian J. Petrol. Geol., Vol. 1, 1992, pp. 75-84.
6. Mathur, S., Sethi, H.S., Goswami, B.G., Pandey, S.C., Sapri, R.K., Chaudhary, D.R., and Misra, K.N., Biomarker geochemistry of oils of Neelam field and some DCS structures, in proceedings of Second Seminar on Petroliferous Basins, India, S.K. Biswas, ed., Dehra Dun, Indian Petroleum Publications, 1993, pp. 443-454.
7. Samanta, U., and Kumar, A., Source rock geochemistry and habitat of oil in the deep continental shelf and shelf margin depression of the Bombay offshore basin, India, in Proceedings of Second Seminar on Petroliferous Basins, India, S.K. Biswas, ed., Dehra Dun, Indian Petroleum Publications, 1993, pp. 435-441.
8. Plummer, P.S., Origin of beach-stranded tars from source rocks indigenous to Seychelles, AAPG Bull., Vol. 80, 1996, pp. 323-339.
9. Plummer, P.S., Diversity of plays bodes well for promising search off Seychelles, OGJ, Dec. 27, 1993, pp. 105-112.
10. Plummer, P.S., Mesozoic source rocks and hydrocarbon potential of the Seychelles offshore, J. Petrol. Geol., Vol. 17, 1994, pp. 157- 176.
11. Plummer, P.S., and Belle, E.R., Mesozoic tectono-stratigraphic evolution of the Seychelles microcontinent, Sed. Geol., Vol. 96, 1995, pp. 73-91.
12. Lawver, L.A., Coffin, M.F., and Gahagan, L., The Mesozoic break-up of Gondwana, in First Indian Ocean Petroleum Seminar, Proceedings of the U.N. Seminar, Seychelles, P.S. Plummer, ed., 1992, pp. 345-356.
13. Plummer, P., Reef mounds indicate timing of hydrocarbon charge off Seychelles, OGJ, July 6, 1998, pp. 85-88.
14. Biswas, S.K. and Thomas, J., The Deccan traps and Indian Ocean volcanism, in First Indian Ocean Petroleum Seminar, proceedings of the U.N. Seminar, P.S. Plummer, ed., Seychelles, 1992, pp. 187-209.
15. Geotrack International Pty. Ltd. and Waples, D., Seychelles Microcontinent: Constrained basin modelling using apatite fission track analysis and vitrinite reflectance to assess the history of hydrocarbon generation, Geotrack Report No. 622, 1997, 415 pp.
16. Sweeney, J.J., and Burnham, A.K., Evaluation of a simple model of vitrinite reflectance based on chemical kinetics, AAPG Bull., Vol. 74, 1990, pp. 1,559-70.
17. Cook, A.C., Davis, A., David, P., and Depers, A.M., ICCP Accreditation system for petrographic analyses, proceedings of the New Zealand Coal Conference, Wellington, October 1997.
18. Duddy, I.R., Green, P.F., Hegarty, K.A., and Bray, R.

    , Reconstruction of thermal history in basin modelling using apatite fission track analysis: what is really possible?, Offshore Australia Conference Proceedings, Vol. I, 1991, pp. III-49 - III-61.

19. Green, P.F., Laslett, G.M., and Duddy, I.R., Mechanisms and kinetics of apatite fission-track annealing-discussion, Am. Mineralogist, Vol. 78, 1993, pp. 441-445.
20. Green, P.F., Hegarty, K.A., and Duddy, I.R., Timing as a vital element of the petroleum system: use of thermal history reconstruction using AFTA and VR as a route to more efficient exploration, Indonesian Petroleum Association, Proceedings of the Petroleum Systems of SE Asia and Australasia Conference, May 1997, pp. 899-912.
21. Morgan, W.J., Hotspot tracks and the early rifting of the Atlantic, Tectonophys., Vol. 94, 1983, pp. 123-139.

The Authors