Exxonmobil Upstream Research Company

Brian P West

John E Eastwood

Steven R May

Michael D Gross

Dominique Gillard

David H Johnston

Dwight C Dawson

Spyridon K Lazaratos

Yao-Chou Cheng

Stefan Hussenoeder

Kevin M Bohacs

John A Farre

George J Grabowski

Thomas J Frantes

Alan E Schwartzbard

Theodore G Apotria

Peter Varnai

R David Potter

Kim Zauderer

A method for retrodicting source-rock quality and/or paleoenvironmental conditions are disclosed. A first set of system variables associated with source-rock quality is selected (). A second set of system variables directly or indirectly causally related to the first set of variables is also selected (). Data for variables selected to be known quantities are estimated or obtained (). A network with nodes including both sets of variables is formed (). The network has directional links connecting interdependent nodes (). The directional links preferably honor known causality relations. A Bayesian network algorithm is used with the data to solve the network for the unknown variables and their associated uncertainties (). The variables selected to be unknowns can be input nodes (paleoenvironmental conditions), intermediate nodes, output nodes (source rock quality), or any combination thereof.

Retrodicting source-rock quality and paleoenvironmental conditions

A method of analyzing dip in a seismic data volume in which a horizontal gradient is calculated in a first direction in the seismic data volume. A vertical gradient is calculated at data locations in the seismic data volume corresponding to the locations at which the horizontal gradient was calculated. Dip is calculated in the first direction from the horizontal gradient in the first direction and the vertical gradient. Repetition of the process for the entire seismic data volume results in a dip volume.

Method for analyzing dip in seismic data volumes

A method of calculating a throw volume corresponding to a seismic data volume. A range of time shifts and a search direction for the seismic data volume are selected. A data location separation and a vertical time window are also selected. A cross-correlation is calculated between data values corresponding to first and second data locations separated by the data location separation and symmetrically located in the search direction on each side of a target data location. The cross-correlation is calculated throughout the vertical time window for each time shift in the range of time shifts. The time shift corresponding to the maximum calculated cross-correlation is stored in the throw volume.

Method of calculating a throw volume for quantitative fault analysis

Seismic horizons are transferred between two three-dimensional seismic data volumes which cover the same area. The seismic data volumes are time-aligned to generate a time-shift volume. A seismic horizon is selected and the time-shift volume is applied to the seismic horizon, generating a time-shifted seismic horizon.

Method for automated horizon transfer and alignment through the application of time-shift volumes

A method for assessing the suitability of seismic data for quantitative amplitude analysis, where the concern is excessive residual normal moveout (RNMO). The invention uses a near offset stack and a far offset stack, the time difference between the two, a mute pattern, a reflection shape assumption for the RNMO, and a waveform and frequency for the far stack traces to generate a formula that estimates far stack amplitude error caused by RNMO. The formula can be used to compensate the far stack amplitude where the error is not so great as to require reprocessing of the data. The method can also be applied to interpreted amplitude maps.

Method for post processing compensation of amplitude for misaligned and misstacked offset seismic data

Near-offset and far-offset seismic data volumes are time-aligned by first selecting a plurality of time shifts. The near-offset and far-offset seismic data volumes are cross-correlated at the plurality of time shifts. An initial time-shift volume and a maximum correlation volume are created from the maximal cross-correlations at the plurality of time shifts. Areas of high time shift from the initial time-shift volume and areas of low cross-correlation from the maximum correlation volume are determined. The determined areas of high time shift and low cross-correlation are filtered from the initial time-shift volume, generating a filtered time-shift volume. The filtered time-shift volume is applied to the far-offset seismic volume to generate a time-aligned far-offset volume.

Method for time-aligning multiple offset seismic data volumes

A method of calculating reflection curvature in a seismic data volume wherein an apparent dip value is calculated in a first direction to generate a first apparent dip volume. A horizontal gradient is calculated in the first direction in the first apparent dip volume using a specified length scale to generate a first curvature volume. The process may be repeated one or more times, and the individual curvature volumes combined to generate a combined curvature volume for the seismic data volume.

Method for analyzing reflection curvature in seismic data volumes

AVO anomalies are classified in near-offset and far-offset seismic data volumes, by first calculating a plurality of initial AVO seismic attributes representative of the offset seismic data volumes. A probabilistic neural network is constructed from the calculated initial AVO seismic attributes. AVO anomaly classifications are calculated in a portion of the offset seismic data volumes. The preceding steps are repeated until the calculated AVO anomaly classifications in the portion of the offset seismic data volumes are satisfactory. AVO anomaly classifications are calculated throughout the offset seismic data volumes using the constructed probabilistic neural network.

Method for classifying AVO data using an interpreter-trained neural network

A method for training a probabilistic neural network to map seismic attributes or similar quantities.

Method for mapping seismic attributes using neural networks

Structural and stratigraphic discontinuities are identified in a 3-D volume of seismic data samples, by first selecting a plurality of directions containing a primary direction and at least one secondary direction. Next, one-dimensional, two-trace first discontinuity values are calculated along the primary direction for each seismic data sample in the 3-D data volume. Next, a series of sequentially less restrictive thresholds is defined, such that a significant portion, preferably at least approximately 10%, of the first discontinuity values satisfy the first threshold. This significant portion of first discontinuity values is then stored in an output discontinuity volume at the corresponding sample locations. The following steps are repeated for each remaining data sample until that sample has a value stored at the corresponding sample location in the output discontinuity volume. First, one-dimensional, two-trace discontinuity values are calculated for the sample sequentially along the secondary directions and the first of the calculated discontinuity values satisfying the first threshold is stored in the output volume. If none of the calculated discontinuity values satisfies the first threshold, then, the next less restrictive threshold in the series of thresholds is selected. Finally, the discontinuity values calculated along the primary and secondary directions are compared sequentially to the selected threshold and the first of the discontinuity values that satisfies the selected threshold is stored in the output discontinuity volume.

Method for imaging discontinuites in seismic data

Seismic facies are identified in a volume of seismic data, wherein, first, a plurality of initial textural attributes representative of the volume of seismic data are calculated. Next, a probabilistic neural network is constructed from the calculated initial textural attributes. Then, final textural attributes are calculated throughout the volume of seismic data. Finally, the calculated final textural attributes are classified using the constructed probabilistic neural network.

Growing community of inventors

Houston, TX, United States of America

Brian P West