Method and apparatus for delineating bed boundaries in subsurface formations and for producing indications of the angle of dip thereof

• US 5,299,128 A
• Filed: 10/05/1990
• Issued: 03/29/1994
• Est. Priority Date: 10/05/1990
• Status: Expired due to Term

• Current Assignee: Schlumberger Technology Corporation (Schlumberger NV)
• Original Assignee: Schlumberger Technology Corporation (Schlumberger NV)
• Inventors: Ganem, Herve, Antoine, Jean-Noel, Delhomme, Jean-Pierre
• Primary Examiner(s): Envall, Jr., Roy N.
• Assistant Examiner(s): BODENDORF, ANDREW
• Application Number:
  US07/593,109
• Time in Patent Office: 1,271 Days
• Field of Search: 364/421, 364/422, 73/152, 73/153, 367/68, 367/69, 367/38, 367/25, 367/28
• US Class Current: 702/10
• CPC Class Codes: G01V 11/00 Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00
• CPC Class Codes: G01V 3/38 Processing data, e.g. for analysis, for interpretation, for correction

Abstract

Data representing a two-dimensional image of the resistivity of a sector of a borehole wall is processed to produce selected flow-line data delineating the boundaries of beds of subsurface earth formations along sectors of the borehole. The latter bed-delineating image data for several circumferentially-spaced sectors is further processed to derive and present data defining the orientations of dip planes, by matching the bed-delineating flow-lines for the several sectors. This is done by focusing the latter flow-lines on a root flow-line, correlating those flow-lines to the root flow-line, and linking the flow-lines to the root flow-line according to a path of minimum cumulative cost.

Claims

1. A method of delineating boundaries of beds in an earth formation traversed by a borehole, comprising the steps of:

(a) producing from original image data containing information about a physical parameter of the earth formation a data field of flow-lines corresponding to directions of least change of the parameter in the image wherein the original image data is representative of variations of the physical parameter of the formation along the longitudinal direction of the boreholes and along a circumferential segment of the borehole wall, which respectively correspond to the longitudinal and the lateral dimensions of said image; and

(b) selecting from the flow-line data field thus produced data corresponding to a subset of said flow-lines which delineate bed boundaries wherein said subset of flow-lines contains data corresponding to dip of the beds.

2. The method of claim 1, wherein the selecting step comprises the steps of deriving, from data as to each flow-line of said field, a characteristic value of said parameter for the corresponding respective flow-line thereby creating a set of characteristic values, and selecting, from within the set of characteristic values thus derived, those values corresponding to bed boundaries.

3. The method of claim 2, wherein said characteristic value is the median value of said parameter in the data for each flow-line.

4. The method of claim 2 or claim 3, wherein said values corresponding to bed boundaries are identified by determining inflection points in a curve consisting of the characteristic values for the flow-lines of said field.

5. The method of claim 4, comprising the steps of determining for each inflection point a significance value, comparing this significance value to a predetermined threshold, and selecting only the inflection points having a significance value above said threshold to identify bed boundaries.

6. The method of claim 1, wherein the step of producing a data field of flow-lines comprises the steps of producing from said original image data, consisting of a two-dimensional function I(x,y) where w and y are coordinates of each image point respectively in the longitudinal and the lateral dimensions, data respresenting a field of gradient vectors;

deriving, from said gradient-vector representing data, data representing a set of tangent vectors which are orthogonal to the gradient vectors; and

deriving from said set of tangent-vector data a set of data representing a field of flow-lines having as their tangent at every point said tangent values.

7. The method of claim 6, wherein said field of gradient vectors is produced by applying a convolution mask to the original image data.

8. The method of claim 6 or claim 7 comprising, prior to deriving the tangent vectors, the step of filtering the field of gradient vector data by squaring the gradient vectors and Gaussian filtering the squared gradient vectors.

9. A method of delineating bed boundaries of earth formations traversed by a borehole, comprising the steps of:

(a) producing data representative of a field of flow-lines in an image corresponding to minimal anisotropy directions of the flow-lines in the image, said data being derived from first image data which are representative of variations of a physical parameter of the earth formations along the longitudinal direction of the borehole and along a first circumferential segment of the borehole wall, which respectively correspond to the longitudinal and the lateral dimensions of the image,

(b) selecting from the data field thus produced a first subset of flow-line data corresponding to bed boundaries wherein said subset of flow-line data contains information concerning dip of the beds,

(c) repeating steps (a) and (b) with at least one further set of image data corresponding to a further circumferential segment of the borehole wall angularly spaced apart from the first segment, whereby data from a further subset of flow-lines corresponding to bed boundaries is selected, and

(d) combining the subsets of flow-lines obtained as a result of steps (b) and (c) so as to match the flow-lines of respective subsets which correspond to the same bed boundaries, thereby generating set of connected flow-lines with each set defining a bed boundary.

10. The method of claim 9, further comprising the step of producing a geometrical description of bed boundaries from said sets of connected flow-lines obtained as a result of step (d).

11. The method of claim 10, wherein the selecting step comprises the steps of deriving, for each flow-line of said field, a characteristic value from the data forming the respective flow-line, and identifying, within the characteristic values thus derived, values corresponding to the bed boundaries.

12. The method of claim 11, wherein said characteristic value is defined as the median of the data for each flow-line.

13. The method of claim 12, wherein the values corresponding to bed boundaries are identified by forming a curve consisting of the characteristic values obtained for the flow-lines of said field, and determining inflection points in said curve.

14. The method of claim 13, further comprising the steps of determining for each inflection point a significance value, comparing this significance value to a predetermined threshold, and selecting only the inflection points having a significance value above said threshold to identify bed boundaries.

15. The method of claim 10, wherein the step of producing a field of flow-lines comprises the steps of producing from said image, which is a two-dimensional function I(x,y), where x and y are the coordinate of each image point respectively in the longitudinal and the lateral dimensions, a field of gradient vectors, deriving a set of tangent values orthogonal to the gradient vectors thus produced, and constructing from said set of tangent values a field of flow-lines having as their tangent at every point said tangent values.

16. The method of claim 15, wherein said field of gradient vectors is produced by applying a convolution mask to the image.

17. The method of claim 16, further comprising the step of filtering the field of gradient vectors by squaring the gradient vectors and Gaussian filtering the squared gradient vectors prior to deriving the set of tangent values.

18. The method of claim 10, wherein said description is a dipmeter plot.

19. The method of claim 9, wherein step (d) of combining the subsets of flow-lines comprises the steps of:

(d1) forming a first set of flow-lines by selecting for each flow-line in the first subset, one flow-line in the further subset, and denoting said flow-line from the first subset a root flow-line the selection being based on at least one criterion indicative of the likelihood that said first set of flow-lines shows the same bed boundary,

(d2) repeating step (d1) to form other sets of flow-lines by successively taking as the root flow-line the flow-lines selected as a result of step (d1),

(d3) retaining said first set of flow-lines as a set of connected flow-lines where the sets formed as a result of steps (d2) are identical with said first set, and

(d4) selecting as a set of connected flow-lines, among the sets resulting from steps (d1) and (d2), the set which optimizes a likelihood to show the same bed boundary where one of the sets formed as a result of set (d2) is different from said first set.

20. The method of claim 19, wherein step (d1) comprises the steps of determining an average direction of the flow-lines of all the images within a predetermined window centered at said root flow-line, and selecting in said further subset the flow-line which is closest to said average direction.

21. The method of claim 19, wherein step (d1) comprises the step of selecting in the further subset the flow-line for which the variations of said physical parameter along the longitudinal direction within a predetermined window is most similar to the variations of the physical parameter shown by the root flow-line.

22. The method of claim 19, wherein step (d1) comprises the steps of:

(i) determining an average direction of the flow-lines of all the images within a predetermined window centered at said root flow-line, and selecting in each of said other subsets the flow-line which is closest to said average direction,

(ii) selecting in each of the other subsets the flow-line for which the variations of said physical parameter along the longitudinal direction, within a predetermined window, are most similar to that of the root flow-line,

(iii) retaining the set of flow-lines resulting from step (i) when it is identical with the set resulting from step (ii),

(iv) selecting among said sets of flow-lines the set to be retained by applying a test comprising a combination of at least the test of step (i) and the test of step (ii) if the sets resulting from step (i) and step (ii) are different.