Lithology identification is a fundamental task in reservoir characterization. Geologists
usually perform this task in core analysis, but core acquisition in a oil field is made in
a small number of wells and in many instances, it is not feasible, as in horizontal
wells. The wireline logging, in contrast with coring process, it is done in all wells,
supplying continuous information along the well depth that may be used for lithology
identification. The objective of this work is to present an intelligent algorithm based
on fuzzy inference for patterns recognition and transformation of log readings in
information about lithology. This method uses of a conventional group of logs
(gamma ray, sonic, density and neutron porosity) and a fuzzy calibration in a cored
well for lithology identification. An application is presented using boreholes from
Namorado Oil Field, in Campos' Basin.
|
