Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Earth science
Reexamination Certificate
2002-09-10
2004-06-29
McElheny, Jr., Donald E. (Department: 2887)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Earth science
C702S016000
Reexamination Certificate
active
06757615
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the field of seismic data processing. Specifically, the invention is a method for transferring seismic horizon interpretations between three-dimensional volumes.
2. Description of the Related Art
An important trend in the petroleum exploration and production industry is the increased desire to rely on seismic data to guide appraisal and development following discovery. This, in turn, results in the generation of a greater number of versions of the seismic data volumes that result from the acquisition and processing of seismic data. For example, at least one, and often several, seismic surveys are required for exploration and prospect delineation. The data from each of these surveys may be interpreted in several different offset stack seismic volumes, and possibly reprocessed for specification of impedance and phase volumes. For the development and production phases, further seismic data acquisition is often needed, for example to acquire higher frequency data or for seismic time-lapse reservoir monitoring.
A common, virtually unavoidable, consequence of the seismic interpretation process is the shifting of the location of horizons between a reference survey and the different vintages of seismic data volumes that result from reprocessing later seismic surveys, or between 3D and 2D seismic surveys within the same area. The process of horizon transfer and alignment is often tedious, and the various techniques that are presently used, such as applying seismic mis-ties to horizons and horizon shifting and snapping, often result in an unsatisfactory correlation between the shifted horizons and the reference surveys. More specifically, a problem that must be addressed is the vertical time variance of the shift needed to align horizons along each seismic line. That problem can make single shifts inaccurate for multiple horizons or even for single horizons covering large areas.
One approach that has been employed in industry is simply to output the original interpretation and import it into the new volume data set. The original interpretation is then used as a guide and the horizon is re-interpreted in the new data set. This approach is computationally inefficient, and the re-interpretation requirement limits its usefulness.
Another approach that has been employed is to interpret a reconnaissance horizon grid and interpolate the grid to create a surface to use as a reference surface for the original horizon requiring a shift. The mis-ties are then computed, usually using a commercial product, and a static shift mis-tie is determined to apply to the original horizon. Limitations of static shift mis-ties constrains this approach. For example, one is that different seismic volumes have dynamic time mis-ties so that the correction of mis-ties for multiple horizons requires replication of the approach, whereas the horizon alignment approach compensates for dynamic time mis-ties for multiple surfaces between different volumes of seismic data applied from a single determination. Gridding of surfaces also introduces errors inherent to the gridding process, while the horizon alignment approach completely avoids gridding of surfaces.
A third common method to align a previously interpreted horizon to a new version of the original seismic data is to copy the horizon to the new seismic data, estimate the shift for the horizon (typically determined through inspection by the interpreter), estimate a snap window for the shifted horizon, and then snap (a procedure that assigns a value at each trace of the surface to a user-specified seismic property or attribute of the trace, such as maximum and minimum values within a user specified time window) the horizon. The accuracy of the results of this procedure is in part dependent upon the snapping parameter choices that are made by the interpreter. The procedure can be accurate when the new seismic data does not significantly vary from the old version. More commonly, however, newly acquired data, and the processing and reprocessing of the original data, results in non-systematic misalignments between the new and reference data, thereby limiting the usefulness of this procedure.
Frequency (bandwidth) variations between the reference volume and the new volume(s) are also a common cause of horizon misalignment. Frequency differences, and differences in the migration velocities used in generating the volumes, often result in non-systematic misalignment. The complexity of this misalignment increases when other factors are added, such as seismic artifacts, different offset angle stacks, and when AVO analysis is performed, such as for Class 1, Class 2, and Class 3 amplitude anomalies. As will be understood to those skilled in the art, AVO amplitude anomalies are classified in terms of the local increase or decrease in reflection amplitude with varying offset angles, such as is caused by the varying impedances of adjacent geologic layers.
Other techniques have been used in the industry to tackle the problem. For example, the commercial seismic interpretation system Geoframe (IESX), of the GeoQuest division of Schlumberger Corporation, includes a MisTie analysis option. In this option mis-ties are calculated using either of two methods. The first method uses a statistical correlation approach between the seismic data at each intersection. The second method measures the mis-tie between interpreted horizons at each intersection. With either method, only a static shift is applied, and only one value per line intersection is permitted, whether it is from a single correlation or from an average of mis-ties from numerous horizons at that intersection. The user can specify whether a static shift is applied for all intersections, applied for some intersections, or a different shift value is applied to different intersections, and the corrections are applied to user-selected horizons.
There are a number of limitations of this approach. First, it is constrained to computation of constant (single static shift for a line) and/or variable (spatially varying) corrections, but cannot apply a dynamic shift to a seismic trace. The approach is therefore best suited for correcting static mis-ties. Second, the alignment for horizon transfer requires two intersecting surfaces to calculate mis-ties. Third, there is no ability to calculate a dynamic alignment along each seismic trace. Fourth, the alignment cannot be calculated between 3D volumes or between 2D seismic lines. Finally, the alignment technique does not utilize a time-shift (or lag) volume or associated correlation (confidence) volume output, which would contain dynamic shifts along each trace and allow, in effect, a volume of alignment corrections to be applied to all horizons and faults. The abstract of D. L. Brumbaugh, “SMAP (Seismic Mistie Adjustment Procedure) Revisited and Revised”, 61st Annual SEG Int. Mtg., Houston, Nov. 10-14, 1991, Expanded Tech Program Abstr. Biogr. V1, pp. 332-334, applies a static shift to seismic lines to match other interpretations and well data. The SMAP revision allows for any orientation of seismic line to be optimally corrected, but does not accommodate dynamic shifts along a trace.
W. L. Walters' U.S. Pat. No. 5,132,938, titled “Adjusting Seismic Data to Tie Other Data”, issued Jul. 21, 1992, generates sets of data that are arranged according to the x-y-z coordinates of seismic lines or as isolated x-y-z points, such as well data. Different sets of data can be compared for the time gate about a common subsurface feature. Then a time delay is determined for each of the trace pairs from the different data sets. As will be understood to those skilled in the art, the terms time gate and time delay refer generally to interpretational differences which are more commonly referred to as mis-ties. These time delays are corrected to the reference surface using a least squares planar fit. The essential aspect of this method is that it requires a previously defined interpretation, whether seismic horizons or
Eastwood John E.
Potter R. David
West Brian P.
Zauderer Kim
ExxonMobil Upstream Research Company
McElheny Jr. Donald E.
Plummer J. Paul
LandOfFree
Method for automated horizon transfer and alignment through... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for automated horizon transfer and alignment through..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automated horizon transfer and alignment through... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3311314