Method and Apparatus For Analyzing Three-Dimensional Data

• US 20100274543A1
• Filed: 11/13/2008
• Published: 10/28/2010
• Est. Priority Date: 12/21/2007
• Status: Active Grant

• Current Assignee: Kelly G. Walker, Paul A. Dunn, Timothy A. Chartrand
• Original Assignee: Kelly G. Walker, Paul A. Dunn, Timothy A. Chartrand
• Inventors: Chartrand, Timothy A., Dunn, Paul A., Walker, Kelly G.
• Granted Patent: US 8,370,122 B2
• Time in Patent Office: Days
• US Class Current: 703/6
• CPC Class Codes: G01V 1/302 in 3D data cubes
• CPC Class Codes: G01V 2210/641 Continuity of geobodies
• CPC Class Codes: G01V 2210/644 Connectivity, e.g. for fluid movement
• CPC Class Codes: G01V 99/00 Subject matter not provided for in other groups of this subclass

Abstract

A method and apparatus for segmenting geobodies extracted from three-dimensional geophysical survey data volumes. In one embodiment, the invention involves (a) obtaining a plurality of data elements, respectively corresponding to three-dimensional locations in a subsurface region, and generating a data volume consisting of discrete cells from the plurality of data elements (1); (b) identifying an initial geobody corresponding to a potential hydrocarbon reservoir within the data volume by grouping cells in the data volume according to a selected connectivity criterion (2); (c) selecting a rule (3) that characterizes strati graphically reasonable geobodies (e.g. they cannot vertically overlap); and (d) in response to a determination that the initial geobody does not conform to the rule, segmenting the initial geobody into a plurality of fundamental geobodies via processing the data elements with an automated segmentation routine, wherein the fundamental geobodies each conform to the rule (4).

Claims

1. An oil and gas prospecting method for predicting connectedness between parts of a potential hydrocarbon reservoir identified in a geophysical data volume corresponding to a subsurface region, comprising:

obtaining a plurality of data elements, respectively corresponding to three-dimensional locations in the subsurface region, and generating a data volume consisting of discrete cells from the plurality of data elements;

identifying an initial geobody corresponding to a potential hydrocarbon reservoir within the data volume by grouping cells in the data volume according to a selected connectivity criterion;

selecting a rule that characterizes stratigraphically reasonable geobodies;

in response to a determination that the initial geobody does not conform to the rule, segmenting the initial geobody into a plurality of fundamental geobodies via processing the data elements with an automated segmentation routine, wherein the fundamental geobodies each conform to the rule; and

using the fundamental geobodies to predict connectedness between parts of the potential hydrocarbon reservoir.

2. The method of claim 1, wherein the geophysical data are seismic data or seismic attribute data.

3. The method of claim 1, further comprising after a determination that the initial geobody does not conform to the rule but before segmenting:

selecting a threshold considered to approximate noise levels in the data, then identifying the cells that are responsible for the initial geobody'"'"'s nonconformance to the selected rule, then adjusting the initial geobody by removing any such cells having data values below the selected noise-related threshold, then repeating the determination of whether the initial geobody, after said adjustment, conforms to the rule.

4. The method of claim 1, wherein the rule is a lateral bifurcation criterion, whereby a criterion is selected limiting geobody branching in a horizontal dimension.

5. The method of claim 1, wherein the rule is a law of superposition, limiting the extent to which a geobody may overlap itself in the vertical dimension, wherein two cells in a geobody are considered to overlap if they have the same horizontal coordinates and are separated vertically by at least one cell not in the geobody.

6. The method of claim 5, wherein the automated segmentation routine uses a shortest path search algorithm.

7. The method of claim 6, wherein segmenting the initial geobody into a plurality of fundamental geobodies comprises:

(i) identifying all zones of overlapping cells within the initial geobody;

(ii) identifying all cells within the potential geobody where overlap is generated by vertical branching, and designating such cells as barrier cells for shortest path searches in subsequent steps;

(iii) for each pair of cells lying one each in overlapping zones, using a shortest path search algorithm to find the shortest path between the pair of cells, and finding the cell of steepest slope for each shortest path and designating it as a barrier cell;

(iv) repeating step (iii), observing barrier cells, until there are no longer any paths between any overlapping zones; and

(v) segmenting the initial geobody into two or more fundamental geobodies using the designated barrier cells as boundaries.

8. The method of claim 6, wherein the shortest path search algorithm is the A* algorithm.

9. The method of claim 5, wherein the automated segmentation routine uses a 3D network of boundaries developed by performing statistical clustering analysis on the initial geobody.

10. The method of claim 9, wherein the automated segmentation routine uses a shortest path search algorithm to find shortest paths between pairs of at least partially overlapping clusters within the initial geobody.

11. The method of claim 9, wherein the geophysical data are seismic data or seismic attribute data, and segmenting the initial geobody into a plurality of fundamental geobodies comprises:

clustering cells in the potential geobody based on positional data or seismic attribute data or a combination thereof, forming a plurality of non-overlapping clusters which are identified as fundamental geobodies.

12. The method of claim 11, wherein segmenting the initial geobody into a plurality of fundamental geobodies further comprises:

(i) performing statistical clustering using one or more selected spatial or attribute criteria until no individual cluster violates the selected rule for stratigraphic reasonableness;

(ii) identifying number and spatial locations of all clusters using one or more selected connectivity criteria;

(iii) detecting all possible pairs of clusters that do not conform to the selected rule for stratigraphical reasonableness;

(iv) selecting a pair of clusters that do not conform to the rule and applying an A* shortest path algorithm to identify a shortest path between the selected pair of clusters;

(v) identifying a location of maximum change where the shortest path crosses a boundary between adjacent clusters and designating all cells having the same horizontal coordinates as that location as a barrier cell set throughout the thickness of the initial geobody;

(vi) repeating steps (iv) and (v) until there are no longer any possible paths between the selected pair of clusters;

(vii) repeating steps (iii) through (vi) until there are no longer any cluster pairs that violate the selected rule; and

(vii) segmenting the initial geobody into two or more fundamental geobodies using the designated barrier cells.

13. The method of claim 9, wherein the statistical clustering analysis is k-means clustering or hierarchical divisive clustering.

14. The method of claim 9, wherein the statistical clustering analysis is hierarchical agglomerative clustering.

15. The method of claim 1, further comprising planning development of the potential hydrocarbon reservoir using connectivity assumptions based on the fundamental geobodies.

16. An oil and gas prospecting method for predicting connectedness between parts of a potential hydrocarbon reservoir identified in a geophysical data volume corresponding to a subsurface region, comprising:

(a) obtaining a seismic or seismic attribute data volume of discrete cells, said data volume representing the subsurface region, and identifying an initial geobody corresponding to a potential hydrocarbon reservoir within the data volume by grouping cells in the data volume according to a selected connectivity criterion based on data changes from cell to cell;

(b) in response to a determination that the initial geobody contains vertically overlapping regions, finding the shortest path between a selected pair of vertically overlapping regions;

(c) breaking the shortest path at its steepest vertical gradient by creating barrier cells or by removal of cells from the initial geobody;

(d) repeating (b) and (c) honoring the barriers until no path exists between the selected pair of overlapping regions;

(e) repeating (b) through (d) for all remaining pairs of vertically overlapping regions;

(f) segmenting the initial geobody into fundamental geobodies according to barrier cells and removed cells; and

(g) using the fundamental geobodies to predict connectedness between parts of the potential hydrocarbon reservoir.

17. The method of claim 16, wherein the shortest path is found using the A* search algorithm.

18. A method of producing hydrocarbons from a reservoir in a subsurface region, comprising:

(a) obtaining a seismic or seismic attribute data volume representing the subsurface region, wherein an initial geobody is identified, using an assumed connectivity criterion, as corresponding to a potential hydrocarbon reservoir, said initial geobody containing portions overlapping other portions in the vertical dimension;

(b) obtaining a processed version of the data volume wherein the initial geobody is segmented into two or more fundamental, non-self-overlapping geobodies, said segmentation having been performed by steps comprising;

(i) using a shortest path search algorithm, finding all possible shortest paths between portions of vertically overlapping regions in the initial geobody;

(ii) breaking the shortest paths by creating barrier cells or by removal of cells from the initial geobody;

(iii) segmenting the initial geobody into fundamental geobodies according to barrier cells and removed cells;

(c) using the fundamental geobodies to predict connectedness between parts of the potential hydrocarbon reservoir; and

(d) producing hydrocarbons from the subsurface region using the connectedness predictions to plan reservoir development.

19. An oil and gas prospecting method for predicting connectedness between parts of a potential hydrocarbon reservoir identified in a geophysical data volume corresponding to a subsurface region, comprising:

representing the subsurface region as a plurality of voxels;

associating a respective data sample with each of the plurality of voxels;

identifying an initial geobody in the subsurface region via aggregating selected ones of the plurality of voxels;

segmenting the initial geobody into a plurality of fundamental geobodies via hierarchical processing; and

using the fundamental geobodies to predict connectedness between parts of the potential hydrocarbon reservoir.

20. The method of claim 19, wherein segmenting the initial geobody into the plurality of fundamental geobodies bodies via hierarchical processing comprises:

(a) initially treating each voxel as a cluster;

(b) selecting a first cluster;

(c) selecting a cluster distance metric;

(d) determining the closest cluster to the first cluster, using the distance metric, and merging that cluster with the first cluster to form a single cluster; and

(e) repeating steps (b)-(d) until all voxels belong to a single cluster.

21. The method of claim 20, wherein segmenting the initial geobody into the plurality of fundamental geobodies via hierarchical processing further comprises recording the clusters merged at each step (d) and the distance between them as determined by the cluster distance metric.

22. The method of claim 21, wherein each of the determined distances has at least four dimensional components.

23. The method of claim 21, wherein each of the determined distances depends upon a parameter derived from a signal that has interacted with the subsurface region.

24. An oil and gas prospecting system for predicting connectedness between parts of a potential hydrocarbon reservoir identified in a geophysical data volume corresponding to a subsurface region, comprising:

means for representing the subsurface region as a plurality of voxels;

means for associating a respective data sample with each of the plurality of voxels;

means for identifying an initial geobody in the subsurface region via aggregating selected ones of the plurality of voxels;

means for segmenting the initial geobody into a plurality of fundamental geobodies via hierarchical processing; and

means for using the fundamental geobodies to predict connectedness between parts of the potential hydrocarbon reservoir.

25. The system of claim 24, wherein the means for segmenting the initial geobody into the plurality of fundamental geobodies comprises a means for attributing a level of overlap of the initial geobody to noise.