Show simple item record

dc.creatorToinet, Sylvainen_US
dc.date.accessioned2012-06-07T23:09:37Z
dc.date.available2012-06-07T23:09:37Z
dc.date.created2001en_US
dc.date.issued2001
dc.identifier.urihttp://hdl.handle.net/1969.1/ETD-TAMU-2001-THESIS-T65en_US
dc.descriptionDue to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.en_US
dc.descriptionIncludes bibliographical references (leaves 121-125).en_US
dc.descriptionIssued also on microfiche from Lange Micrographics.en_US
dc.description.abstractThis study reports an investigation of the potentialities of artificial neural networks in the field of reservoir characterization. A first step has been the review of theoretical principles involved in neural networks computations, in order to select among numerous networks which one could be the most efficient for approximating a relationship between seismic attributes and reservoir parameters (porosity, shaliness, water saturation) derived from well log data. The selected network is trained for learning a relationship between seismic attributes, interpolated at the well locations and reservoir parameters at the same locations. Then, the relationship learned along the well will be applied to the entire volume of a seismic cube to predict reservoir properties at the seismic scale. It turns out that feed-forward networks, trained using a Levenberg-Marquardt algorithm are the most suited to the objective. A method to estimate the accuracy of the predictions has been developed based on the distributions of the values of the seismic attributes and those of the reservoir parameters in the training set, so that probability cubes can be associated with the prediction cubes. Neural networks suited to the predictions of reservoir parameters have been designed and validated using a data set for which the predictions to obtained were known. They have demonstrated excellent prediction abilities. Then neural networks have been applied on a real data set from Edna (Jackson County, Texas). Predictions of porosity, shaliness and water saturation have been performed at the seismic scale using cubes of attributes. The results have been associated with probability cubes that allow the quantification of the accuracy of the predictions and that give pertinent results. A simultaneous study of the predicted cubes of porosity, shaliness, and water saturation, along with their associated probability cubes helped characterizing an already producing reservoir. The three predicted parameters show an excellent correlation with well logs (SP, ILD, and DT).en_US
dc.format.mediumelectronicen_US
dc.format.mimetypeapplication/pdfen_US
dc.language.isoen_USen_US
dc.publisherTexas A&M Universityen_US
dc.rightsThis thesis was part of a retrospective digitization project authorized by the Texas A&M University Libraries in 2008. Copyright remains vested with the author(s). It is the user's responsibility to secure permission from the copyright holder(s) for re-use of the work beyond the provision of Fair Use.en_US
dc.subjectgeophysics.en_US
dc.subjectMajor geophysics.en_US
dc.titleReservoir characterization using seismic attributes, well data, and artificial neural networksen_US
dc.typeThesisen_US
thesis.degree.disciplinegeophysicsen_US
thesis.degree.nameM.S.en_US
thesis.degree.levelMastersen_US
dc.type.genrethesis
dc.type.materialtexten_US
dc.format.digitalOriginreformatted digitalen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

This item and its contents are restricted. If this is your thesis or dissertation, you can make it open-access. This will allow all visitors to view the contents of the thesis.

Request Open Access