--> Seismically-driven characterization of deep-water turbidite systems

AAPG Geoscience Technology Workshop

Datapages, Inc.Print this page

Seismically-driven characterization of deep-water turbidite systems


Turbidite systems or submarine fans are considered to be the most important clastic accumulations in the deep sea. They represent a sediment-transfer system between the hinterland source area and the deep-sea depositional sink. Petroleum exploration in deep-water settings is resulting in the discovery of many giant fields linked to those systems. In the light of high exploration and development costs, an accurately characterize these challenging reservoirs is imperative. In this research, a set of seismically-driven approaches have been chosen to characterize the deep-water turbidite system of the study area. Within and outside the research area, several small hydrocarbon discoveries have been made. Especially the gas and condensate discoveries made in recent years brought further attention to the turbidite reservoir characterization. The research area is covered by 3D high resolution seismic and 12 wells. Seismic well tie has been performed on several wells some of which penetrating the turbidite reservoir formation. This approach allows to study the seismic signature of the turbidite and compare it with the neighboring lithology. Also, it is important to understand the influence of the turbidite thickness on the seismic signature. The extracted wavelets derived from the seismic traces and the log reflectivity were used for 2D modeling that allowed to study the amplitude behavior of the turbidite as a function of its thickness. Fluid substitution gave further insight into the amplitude behavior. RMS amplitude extraction and topographic time mapping can assist in identifying the turbidite accumulations and their depositional routes. However, this approach does not take into account the seismic signature of the turbidite: different signatures may deliver similar RMS amplitudes. Therefore, a waveform classification based on neural network was applied to a time window around the seismic horizon showing the amplitude anomalies linked to the turbidites (Fig. 1). It turned out that a class of high amplitudes showed a different signature compared to the turbidite classes. An inspection of the log data of one well penetrating this area suggests a carbonate zone. This integrated approach based on seismic attributes, waveform classification, 2D modeling and seismic well tie helps in proposing a depositional modal for the deep-water turbidite reservoir of the study area.