--> ABSTRACT: Stochastic Modelling of Turbidite Reservoirs Constrained by Seismic Informations (Campos Basin, Offshore Brazil), by Olinto G. Souza Jr., Remi Eschard, Paulo R. S. Johann, Frederique Fournier, and Lin Y. Hu; #91019 (1996)

Datapages, Inc.Print this page

Stochastic Modelling of Turbidite Reservoirs Constrained by Seismic Informations (Campos Basin, Offshore Brazil)

Olinto G. Souza Jr., Remi Eschard, Paulo R. S. Johann, Frederique Fournier, and Lin Y. Hu

A new methodology of reservoir probabilistic modelling was applied at an oil field scale in the brazilian continental margin. The Namorado field, located in the Campos basin in the passive margin of eastern Brazil, is a mature oil field with 60 wells and 1326 km of 3D seismic lines. The reservoir (Albian/Cenomanian), deposited in an overall trangressive context, produces from turbidite sandstones and is composed by massive sandstones, classical Bouma turbidites, debris-flows, marls and shales. Three depositional sequences were correlated across the field, each of them corresponding to cycles of relative sea-level falls.

The stochastic simulation of the spatial lithofacies distribution as a non stationary truncated gaussian function was carried out to generate realistic reservoir images. The important lateral variation in the facies and thickness of the sedimentary bodies is characteristic of non stationarity. In this case, the seismic data integration is crucial for a successful modelling of the phenomenon. A reservoir oriented processing of the seismic data was carried out, including a post-stack stratigraphic inversion. Seismic attributes on inversion results were calibrated at wells in terms of lithofacies.

In the non stationary approach, the images of the reservoir lithofacies distribution are conditioned to wells and deduced from the Vertical Proportion Matrix, which represents the percentage of each lithofacies at a given spatial coordinate and in each depositional sequence. This matrix, defined over a regular grid, is computed from wells and seismic derived constraints.

The incorporation of the seismic information in the non stationary gaussian function model has ensured a successful reproduction of the reservoir architecture.

AAPG Search and Discover Article #91019©1996 AAPG Convention and Exhibition 19-22 May 1996, San Diego, California