Dabbouk, Chawki¹, Hamad Bu Al-Rougha¹, Hedhili B. Mohamed Gossa¹, Paolo Ferraris², Muatasam H. Al-Raisi²
(1) ZADCO, Abu-Dhabi, United Arab Emirates
(2) SCHLUMBERGER Middle East, Abu-Dhabi, United Arab Emirates

ABSTRACT: Integrated Log Analysis for Facies Identification and Permeability Prediction in a Complex Carbonate Giant Field

Obtaining accurate rock-type descriptions and successful permeability predictions in carbonate reservoirs from logs only is in general a very delicate task. This challenge was addressed using a thorough data-set consisting of logs acquired on six wells drilled on a giant Middle East carbonate field.

Five main factors differentiate the case study described here from what has been attempted in the past:

1) The quality and the diversity of input logs. New technology measurements such as nuclear magnetic resonance (NMR), resistivity images, Stoneley sonic, in addition to the traditional formation evaluation measurements were used to better describe pore sizes and rock texture.
2) The integrated approach target was to compute one optimal descriptive answer combining all the information available from the different logs rather than merging graphically anwers from individual sensors. Neural network analysis was selected as integration tool because an analytical solution was not suitable for this type of environment.
3) A comparison of log-derived properties and core-measured properties was used to find and test correlations between optimal evaluation from logs and ground truth from cores. Core plugs permeability measurements and continuous core description were first applied to identify and train petrophysical and geological relationships using a limited data set (two wells out of the six available) and subsequently the remaining data were applied to test the result obtained.
4) NMR laboratory measurements performed under a four-step differential pressure depletion scheme were the foundation for a comprehensive rock facies data-base model to predict fluid producibility
5) In the course of the study the training set was expanded to include additional core data one well at a time, to fine tune and test spacial dependency of petrophysical properties throughout the field.

This study significantly contributed to:
a) an improved understanding of the applications and limitations of replacing coring with logging
b) the development of an integrated framework for future evaluations

AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.