Seismic Attributes for Lithofacies Prediction Through the Use of Artificial Neural Networks Methods: A Case Study from Bloque VIII Field, Lake Maracaibo, Venezuela

Andres G. Espeso and Ann M. Mezones. PDVSA Intevep / Exploración y Producción, PDVSA Intevep, Urb. Sta. Rosa, Sector El Tambor. Edif. Fase A. Piso 3, Los Teques, 1201, Venezuela, phone: 58-212-330.74.59, fax: 58-212-330.76.33, espesoa@pdvsa.com

To identify and to predict from 3D seismic the lateral distribution of lithofacies in areas where no well information is available, an artificial neural network study was conducted in the C Eocene sandstones of the Bloque VIII field.

There is a large number of oil traps associated to sedimentary discontinuos sand bodies which produces a high stratigraphic complexity to the reservoirs in this field. Areal extention of the sand bodies in the C sandstone were determined from representative crossplots obtained between seismic amplitude attributes and well log data by training of the neural network. The application of this methodology to the Bloque VIII field has improved the understanding of the depositional enviroment by efficiently mapping these discontinuos stratigraphic sand bodies which reduces the risk of new wells, and allows identification of additional not drained plays in the development areas of this field.

Geologic Interpretation of Seismic Attributes and Amplitude Anomalies
2004 AAPG International Conference and Exhibition Technical Program