--> Reservoir 3-D Static Modelling Using Multi-Attribute Seismic Facies Characterization: Example of a Carbonate Lacustrine System From the Kwanza Basin of Angola

AAPG ACE 2018

Datapages, Inc.Print this page

Reservoir 3-D Static Modelling Using Multi-Attribute Seismic Facies Characterization: Example of a Carbonate Lacustrine System From the Kwanza Basin of Angola

Abstract

This work illustrates the construction of a static model during the Exploration phase of the pre-salt carbonate lacustrine play in the Kwanza basin of Angola. At this stage, the aim was to run reservoir simulations in order to optimize the development scenarios and better estimates MEFS and the final economic viability.

Building a static model and predicting spatial distribution of petrophysical properties within such a heterogeneous reservoir at the exploration stage is a real challenge due to the lack of direct information from well data in the area of interest. However, integration of basin scale knowledge such as tectonic evolution, carbonate depositional model, well database including petrophysical properties with 3D seismic data interpretation, can significantly help to achieve a geologically reliable reservoir model.

An integrated approach for facies modeling based on a carbonate depositional model, seismic stratigraphy, Multi-Attribute Seismic Facies Characterization and artificial neural network, was implemented. It was aimed specifically to model petrophysical properties for each facies within a carbonate build-up developed over a basement high in a lacustrine setting according to a described behavior in the nature.

The information obtained from fields located in the same tectonic domain of the hyper-extended rifted margin of the Kwanza Basin was analyzed in detail to generate a regional sequence stratigraphy and a depositional model that would serve as an analog for our area of interest.

Petrographic, sedimentary and stratigraphic information from cores and logs correlation from available Kwanza basin pre-salt wells were crucial to develop a depositional model and facies classification. They were compared to a seismic facies catalog in order to identify the same facies on our 3D seismic data using seismic-stratigraphy, multi-attribute seismic facies characterization and neural network classification.

Finally, each seismic facies could then be related to a sedimentary facies previously defined in the regional model.

The Kwanza basin well database was used to estimate the petrophysical properties range for each specific facies. The reservoir model was then populated within a geostatistical framework to force a natural distribution by using the facies as the main constrain.

The reservoir model was finally run successfully and helped to optimize the development scenarios and reduce the MEFS, obtaining a better economic estimation.