High Resolution Migration Modelling and Seismic Inversion – Example from a Channelized System in Offshore Angola

M. Hertle¹, N. H. Schødt¹, H. P. Hansen¹, and M. De Lind Van Wijngaarden^2
1Maersk Oil & Gas, Copenhagen, Denmark
²IES GmbH, Aachen, Germany

The area evaluated is located inside the Lower Congo Basin, which extends from central to northern offshore Angola. The basin contains an Early Cretaceous clastic rift sequence (including the lacustrine ‘Bucamazi source’), post-rift evaporites (Albian salt) and an up to 3-4 km thick post-salt passive margin sequence. This sequence includes the prolific marine Late Cretaceous to Eocene ‘Iabe and Landana source rocks’ succeeded by the lean mixed gas-oil prone source rocks, the turbidite channel sands and the thick sealing mudstones of the Oligocene to Miocene Malembo Formation. The primary reservoir targets are stacked channel sands of Late Oligocene age. Earliest hydrocarbon expulsion from the Iabe source rock started in the early Miocene and is ongoing up to present day with present day maturities ranging from the main oil generation to gas generation window.

3D hydrocarbon charge modelling technology has advanced significantly in the recent years, especially with the introduction of inversion percolation theory and the integration of advanced PVT calculation of the hydrocarbon phases. One of the main issues in migration modelling is still the actual model building itself, especially regarding the facies (lithology) distribution in more complex geological settings (e.g. channel and amalgamated channel systems).

Seismic inversion methods were used to create a high resolution lithology cube (shale, sand) to be integrated into the petroleum systems model. From well log studies it was found that sand and shale was separable in the relative elastic impedance domain. Using P-sonic, S-sonic and Density log data, angle impedance was derived for the angles 13 degrees and 23 degrees which are comparable to the seismic near- and mid-stack, respectively. Subsequently, the seismic near- and mid-stack were each inverted using a sparse-spike inversion technique to derive seismically based 13 degrees and 23 degrees relative angle impedance volumes. The far-stack was disregarded due to data quality problems. Using these volumes in combination, sand and shale probabilities were then estimated using a supervised-neural-network trained on angle impedances and lithological information at the well location. The final lithology volume was determined by applying probability cut-offs; sand-probability>50%=sand, sand-probability<=50%=shale.

Whereas compaction, pressure, temperature and hydrocarbon expulsion were calculated on a coarse grid, hydrocarbon migration was modelled on a high resolution grid using an inversion percolation simulator. Porosity/effective stress relationships and porosity/capillary entry pressure relationships were developed and tested regarding their sensitivity on the migration results. A significant improvement in the evaluation of the hydrocarbon charge (both migration pathways and accumulated volumes) could be achieved; especially in comparison with the initial coarse migration model and traditional charge analysis.

AAPG Search and Discovery Article #90091©2009 AAPG Hedberg Research Conference, May 3-7, 2009 - Napa, California, U.S.A.