Reservoir Characterization of the Benguela Belize Field, Offshore Angola, using Principle Component Analysis

By

 Joao Nogueira¹, Mark Moon², John Dunn²

(1) Sonangol DPP, Luanda, Angola (2) ChevronTexaco Overseas Petroleum, San Ramon, CA

 The Benguela Belize field is one of a series of sand filled turbidite channel reservoirs that have been discovered in the deep waters of Block 14, off the coast of Angola. With sparse well control and good 3D seismic, characterization of the reservoir for detailed reservoir simulation has focused on the use of seismic attribute processing to help define channel geometries and make rock property predictions.

This poster describes how Principle Component Analysis (PCA), a long established statistical technique used to transform and analyze multivariate data sets, is being used in a new, innovative way to make better predictions of rock properties, and help refine the definition of reservoir geometry and quality. The process also addresses reservoir management issues such as, water-flood sweep efficiency, well positioning, producing guidelines, and completion tactics.

Predictions of Vshale using PCA were made using ChevronTexaco’s PCA proximity transform algorithm. Which in basic terms takes seismic attribute data such as far-offset amplitude, and rearranges it into multi-dimensional principle component space. The method presented combines well data with principle components and uses an inverse distance interpolation to predict rock properties at log resolution scale. In addition, clustering of data within PCA space identifies geobodies which help to improve the imaging of the reservoir’s complex amalgamated channel geometries. By integrating the cluster analysis and proximity transform results into our reservoir model we were able to improve definition of reservoir geometries and provide a good prediction of Vshale from which porosity and permeability properties were populated using cloud transforms.