Generation of a Volume of Reservoir Geological Properties from 3D Seismic and Well Data

Espeso, Andres G.¹, Ann Marie Mezones¹, Asdrubal Bernal¹ (1) PDVSA Intevep, Los Teques, Edo. Miranda, Venezuela

Cores, well logs and seismic data detect different scales of reservoir heterogeneity with different degrees of accuracy and covering. 3D seismic contributes with information in big areas of the field, but it has vertical low resolution, while the well data contribute with precise information about reservoir properties with vertical high resolution in a specific point of the field. Integration of these multi-resolution data sources in reservoir models allows to achieve a better definition of the reservoir internal configuration in its fullest extent.

This study is developed in a field of heavy oil, with Miocene fluvial and deltaic reservoirs located in the western basin of Venezuela.

Well data was analyzed interpreting the geological characteristic of interest to extrapolate it to the inter-well space through seismic data. Several seismic attribute cubes were analyzed to establish the relationship between seismic data and well logs responses.

An estimation model based on artificial neural network was created to assign geological characteristics to the whole seismic volume, obtaining a reservoir geological properties distribution cube of seismic resolution derived from seismic and well log data that represents in a more realistic way the reservoir heterogeneity.

Reservoir geological properties distribution cube allows to know in a detailed form the reservoir, reducing the existent uncertainty on the distribution of their properties in the inter-well space. This cube will be used during the development of the field to optimize the well path planning in new horizontal wells and to design better patterns for the well locations in future injection programs.