AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Noise Into Signal Around the South Atlantic Margins: Reverse Geochemical Analysis Using Spatial Inputs


Seismic processing attempts to separate signal from noise; then attenuate the latter and correctly position the former. Geochemical analysis of oils works in similar fashion to look past maturation, migration and biodegradation effects (noise) to reveal fingerprints in the oil of its source environment(s). We consider spatial parameters such as distance to and depth of depocenters; known and likely barriers to hydrocarbon migration; and clustering of samples versus basin elements to indicate the extent of geochemical “noise” sources. Oils excluded from the final family assignments can still, based on their “noise”, reveal geologic influences on migration and maturation that may also apply to their geochemical neighbours. From the roughly 10% of our sample set in the “noise” category, we present four groups of samples in three basins on the West African and Brazil margins. One example is late gas/condensate stripping and flushing of oils offshore Ivory Coast and Ghana along the West Africa transform margin. Long-distance migration is implied from a highly mature source to updip reservoirs unshielded from late charge. In the Lower Congo Basin of Angola-Cabinda, near-shore samples have been similarly affected. Considering the normal maturity of outboard marine-sourced oils, we conclude that 15 pre-salt reservoired oils (representing one-eighth of the total sample set) were later altered by high-maturity pre-salt sourced hydrocarbons. Of about 380 oils in Brazil's Greater Campos basin, 20% present cases of either overmaturity (38 samples, likely from pre-salt sources) and biodegradation (37 samples of likely pre-salt sourced hydrocarbons) that require separate treatment. We conclude with qualitative rules-of-thumb for reworking existing oils. Newer analytical methods can yield missing information from some samples. Spatial factors reveal or at least strengthen source-reservoir linkages inferred from the oils. Noise is partially converted into signal and our petroleum systems understanding improves by incorporating less reliable samples in our overall analysis.