High-grading blocks with gravity in the Gulf of Guinea

Area of interest and data sets used for gravity interpretation. [29,906 bytes] Euler deconvolution of Bouguer gravity, West Africa region, 0 - 4 km depth slice. [31,039 bytes] Reconstruction of the South Atlantic at 55 Ma (million years before present) displaying horizontal derivative of Bouguer gravity. Gray represents areas where the gravity grid was not originally included (South America) or where crust has been destroyed since 55 Ma. [30,776 bytes]

Maximizing productivity of small staff

William G. Dickson
Dickson International Geosciences
One of the goals of a gravity high-grading project was to allow an independent explo ration and production company with a small staff and budget to evaluate large areas in the Gulf of Guinea. (West Africa). The result ing gravity maps provided a "feel" for the location of source basins and con trols on reservoir distribution, and had impli cations for fluid flow.

This allowed a block ranking by prospectivity to be developed, without expanding the budget and staff. Further work has expanded the area of interest (AOI) with software reconstructions of the South Atlantic and its margin basins.

The project effort was based on gravity but incorporated several other data types to provide quicklook evaluations. Gravity data collected on land, sea, air, and from space were combined with magnetics, SLAR, seismic, geologic data and observations from the literature. The initial AOI was offshore West Africa. With the enthusiasm gained from initial results, the work expanded to the entire South Atlantic and adjacent continents.

The project dataset is a compilation of several datasets at different resolutions. The highest resolution data was on the continental margins and adjacent onshore areas where the main petroleum provinces are located. Data points were on a 4 km by 4 km grid offshore and 11 km by 11 km grid onshore in the striped area. Elsewhere, the onshore grid was from a geoidal model with a resolution of about 50 km by 50 km, while in the oceans, the grid was still 4 km by 4 km, but of lower quality.

The integrated gravity data were upgraded with improved continental margin bathymetry to produce several maps. The initial maps consisted of Total Horizontal Derivative of Bouguer gravity (THD), Isostatic residual gravity (ISO) and Euler deconvolution of Bouguer gravity (EDB).

Euler deconvolution

EDB results were plotted in depth slices of 0-4 km, 4-8 km, 8-10 km, 10-15 km, and 15km to top of source. Each map showed the computed top of a density difference in a given depth range. The stacking of several color-coded slices permitted the identification of the shape and nature of some of the features.

For example, a normal fault will show several solutions (one at each notable density layer) with offsets corresponding to the dip of the bounding fault. Depth slices of EDB solutions behaved in intuitive ways with shallow sources reflecting known tectonic and basin structures while the deeper solutions helped to delineate limits of thinned continental crust.

Near the coast, there were few EDB solutions, mainly over the offshore Congo River. These solutions resulted from computing an inaccurate Bouguer correction for a steep, narrow channel from a coarse grid. The resulting EDB solutions were only on the shallow slices. On deeper slices, they disappeared, implying erosional rather than tectonic control on the canyon. The actual basement trend seen onshore is ENE-WSW. The east-west trend seen offshore on this map is only related to Tertiary lobe-switching, not basement control. Since hydrocarbon migration is affected by the basement trends, this distinction is important.

A second point is that most EDB solutions lie well outboard of the continental-oceanic boundary (COB). In fact, they indicate density contrasts at depths less than the sea floor, which is in the range of 4-5 km. This discrepancy and the randomness of the solution vectors suggest that the EDB solutions are related to noise in the data set. As shown by the striped areas, the satellite data were reprocessed from the shore to beyond the COB, to improve the tie between onshore, marine and satellite data sets. We inferred that the reprocessing also eliminated a significant amount of noise present in the public domain satellite gravity data.

Regional keys

Combinations of gravity products were used regionally to highlight areas for detailed study while minimizing the exploration budget. In combination with a few published seismic profiles and related studies and some very regional non-exclusive seismic lines, gravity maps allowed the determination of three key factors:

  • Location of the COB: The COB mattered because oceanic crust was associated with lower heat flows and deeper waters hence lower sea bed temperature. Thus, much thicker overburden is required for generation of hydrocarbons above oceanic crust than above continental crust.
  • Estimation of the total sedi ment thickness: The total sediment thickness above any postulated source interval must exceed the calculated minimum for hydrocarbon generation. Traps can only be charged if they have access to mature source rocks.
  • Direction of structural trends: The structural trends may not be defined by widely spaced seismic profiles. Tighter seismic grids increase cost and require more interpretive effort. Gravity can clarify these trends, which often control reservoir distribution and hydrocarbon migration.
These keys allowed a robust block classification for hydrocarbon generative potential (with source rocks assumed). Seismic signatures and water depth were factored-in and applied in a block ranking. Based on this preliminary ranking, targeted seismic profiles were licensed to detail just a few blocks. This was a bold move because the area had been almost completely unmapped and there was concern that good prospects would be overlooked without reviewing vast amounts of seismic data.

Results versus costs

The data expenditure was less than 1/4 the cost of all available seismic lines. The staff was too small to evaluate all lines anyway, but the interpreters did concentrate on the lines actually licensed. Their detailed seismic-stratigraphic interpretation with regional impli cations drew on the initial work. Even detailed correlations between seismic lines were strongly influenced by the gravity maps, which showed trends and general locations of highs and lows.

According to several industry scout reports, the final block ranking matched the block popularity. Both management and partners were pleased with how the limited budget and combination techniques were employed to yield a thorough, high quality appraisal.

The West Africa margin THD showed two distinct domains: one full of high amplitude lineaments interpreted as continental, and the other, a quiet zone of oceanic crust. The dividing line was sharp and well defined from the Walvis Ridge to the Niger Delta. One domain was hot (with high amplitudes), dissected and full of linear features corresponding to the continental synrift fabric.

The seaward domain was cool and un disturbed, especially in the magnetically quiet area of Cretaceous crust. Lineaments in the continental crust revealed a synrift pattern of half-graben development. Offsets correlated with known oceanic transforms such as the Fang and N'Komi fracture zones.

Reconstruction

From a confident interpretation on the African side, it was natural to extend the observations to the South American margin. Since the two continents fit at breakup, one passive margin dictated the shape of the other. We used purpose-built plate tectonic reconstruction software, GETplatetrademark, with published plate restoration parameters to re-project the continents and associated data grids for a range of ages.

Instead of typical wireframe reconstructions, this work had grid values at 4 km spacing and maps were produced for 5 ages between 55 Ma (million years before present) and 120 Ma. This provided exploration leverage in two ways. It provided a framework to extrapolate data from known to unknown areas (especially in the South Atlantic synrift basins) increasing the value of data already in hand. And again, it helped target high-potential blocks before big money had to be spent.

The THD fit of the conjugate margins indicated adjustments were needed to the published limits of the synrift Campos and Espirito Santo Basins of Brazil. This resulted in varying the area of inferred lacustrine source potential. This limit was not clearly defined by the THD map on the Brazilian margin, partly due to masking by Tertiary to Recent sedimentation.

The thickness of younger rocks diminished the THD response of the underlying synrift section, requiring both THD and ISO to image synrift and postrift sections. A further masking resulted from the extensive volcanism along the Brazilian margin. Again, the necessary fit to the West African margin helped determine the COB in obscured areas.

Defining an AOI

Another benefit of the reconstructions was the simple visual correlation of large amounts of data. The Cretaceous basins contained the main source rocks for the South Atlantic. Studying them in a present-day configuration meant large maps or small scales and left questions as to the nature of the outboard or distal limits of the basins.

By turning back the clock, the interpreter could see the correct juxtaposition of the basins, visualize their limits, and plot interpretations on a much smaller piece of paper in full color. This made for nearly selfexplanatory maps, excellent tools for partner and management meetings.

The work helped determine factors important to hydrocarbon generation (source rock limits, oceanic versus continental crust, and associated heat flows) and migration, as well as reservoir rock distribution (transforms, fracture zones, and their spatial and temporal persistence).

The costs were modest for the areas covered and in one example, allowed a small team of interpreters to create an excellent product for a fraction of a competitor group's cost.

Acknowledgement

This paper is a condensation of two posters from AAPG Rio '98 augmented with material from SAMBA (South Atlantic Margin Basin Analysis), a non-exclusive report by Dickson International Geosciences and GETECH. William G Dickson can be reached at Tel: 281-556-9212 or Email: billd@hal-pc.org.

Copyright 1999 Oil & Gas Journal. All Rights Reserved.

More in Regional Reports