Stochastic well correlation based on facies and sequence interpretations using the hierarchical algortithm WeCo

Paul Baville and Guillaume Caumon and Marcus Apel and Dirk Knaust and Silvan Hoth and Cedric Carpentier and Christophe Antoine. ( 2019 )
in: 2019 Ring Meeting, ASGA

Abstract

Well-log and core-sample interpretation is a difficult task requiring significant interpretation skills. Thereby interpretation and correlation are intertwined and several interpretationcorrelation loops may be necessary. Consequently, an increasing number of interpreters may generate a larger set of correlations for one and the same reservoir. To reduce interpretation bias due to the interpreter, we propose an automated method to generate stochastic well correlations, which can be scrutinized by experts and used in uncertainty studies. The purpose of this paper is to present new features of WeCo, a tool using the hierarchical method of stochastic multiple well stratigraphic correlation. WeCo utilizes the sedimentological facies correlation based on core-sample interpretations along the well paths with respect to their depositional conditions and in a chronostratigraphic framework. Input data include well data (.las files) and detailed facies interpretation along the well path (as obtained from core data and well logs). These facies type are classified by their depositional environments and in particular by their position along a proximal-to-distal transect. Assuming that wells have a global distality due to their position in the whole basin, these two distalities (global and local) can be compared and used to compute a correlation cost between two markers; i.e. a distal facies in a proximal well cannot be associated with a proximal facies in a distal well. The second assumption is that parasequence boundaries are time lines and can be used to validate the correlations of markers. Assuming that a facies succession is a regressive sequence based on a shallowing-upward trend and that a facies succession with deepening-upward trend is a transgressive sequence, sequence boundaries can be interpreted in well logs and used as stratigraphic markers.

Download / Links

BibTeX Reference

@inproceedings{BavilleRM2019,
 abstract = { Well-log and core-sample interpretation is a difficult task requiring significant interpretation skills. Thereby interpretation and correlation are intertwined and several interpretationcorrelation loops may be necessary. Consequently, an increasing number of interpreters may generate a larger set of correlations for one and the same reservoir. To reduce interpretation bias due to the interpreter, we propose an automated method to generate stochastic well correlations, which can be scrutinized by experts and used in uncertainty studies. The purpose of this paper is to present new features of WeCo, a tool using the hierarchical method of stochastic multiple well stratigraphic correlation. WeCo utilizes the sedimentological facies correlation based on core-sample interpretations along the well paths with respect to their depositional conditions and in a chronostratigraphic framework. Input data include well data (.las files) and detailed facies interpretation along the well path (as obtained from core data and well logs). These facies type are classified by their depositional environments and in particular by their position along a proximal-to-distal transect. Assuming that wells have a global distality due to their position in the whole basin, these two distalities (global and local) can be compared and used to compute a correlation cost between two markers; i.e. a distal facies in a proximal well cannot be associated with a proximal facies in a distal well. The second assumption is that parasequence boundaries are time lines and can be used to validate the correlations of markers. Assuming that a facies succession is a regressive sequence based on a shallowing-upward trend and that a facies succession with deepening-upward trend is a transgressive sequence, sequence boundaries can be interpreted in well logs and used as stratigraphic markers. },
 author = { Baville, Paul AND Caumon, Guillaume AND Apel, Marcus AND Knaust, Dirk AND Hoth, Silvan AND Carpentier, Cedric AND Antoine, Christophe },
 booktitle = { 2019 Ring Meeting },
 publisher = { ASGA },
 title = { Stochastic well correlation based on facies and sequence interpretations using the hierarchical algortithm WeCo },
 year = { 2019 }
}