We present a data-analysis approach to determine Archie parameters m and n from standard resistivity measurements on core samples. The analysis method, core Archie-parameter estimation (CAPE), results in computed water saturations that agree well with core-measured saturations. CAPE determines m and n by minimizing the error between computed and measured water saturations. The conventional method minimizes the error in nonphysical quantities. Also, CAPE provides a natural, physically meaningful method of "averaging" Archie parameters, and with an error statistic, aids in zonation of a well or reservoir into different sets of Archie parameters. Finally, we show that the Archie constant a is a weak-fitting parameter, with no physical significance, that can generally be set to unity.

No abstract

High-resolution biostratigraphic analyses provide calibration of both depositional systems and specific depositional environments. Such studies facilitate prediction of reservoir geometry and the lateral continuity of both reservoir sandstones and sealing mudstones. Data from the offshore Nigeria Oso Field are presented as a case study of one late Miocene deltaic depositional system. Within the Oso Field, marine mudstones are recognized as effective top seals across the entire field in contrast to marginal marine to non-marine mudstones which have more restricted distribution and act as intrafield baffles. Three sequence stratigraphic models have been considered: (1) the first is based on biostratigraphic data integrated with core sedimentology; (2) a second model constructed using core sedimentology; and (3) a third based on interpreted regional seismic reflection profiles and well-log data. Integration of all data shows that the lower Oso Field producing interval is interpreted as a lowstand prograding wedge, and the upper interval as a prograding distal transgressive or alternatively distal highstand system tract.

No abstract

Neural Network-based minimum mean square (MMS), maximum a posteriori (MAP), and maximum likelihood (ML) parameter estimation are considered here. Ihe multilayer perceptron (MLP) is shown to approximate the minimum mean square estimator. Linear transfonns are used to compress data for the purpose of eficient parameter estimation. Raw and transform domain lower b o u h are developed and used as an optimality criteria in a procedure defined to compare various transfomations.