Granted Patent B2

US 12,612,855 · App. 18/254,918 · Granted Apr 28, 2026

Inventors: Ishan Raina (Abu Dhabi, AE); Sylvain Wlodarczyk (Montpellier, FR)

Assignee: SCHLUMBERGER TECHNOLOGY CORPORATION

E21B47/09G01V11/00G06F40/295E21B2200/20G01V2210/643

Abstract

Claims (85)

1 . A method, comprising:

receiving a target well log having a name, a unit, and a description assigned thereto;

identifying a plurality of first entries in a database that have a unit and a measurement type assigned thereto, wherein the units of the plurality of first entries are compatible with the unit of the target well log, and wherein the measurement type of the plurality of first entries comprises one or more first n-grams;

generating a corpus of words from the measurement types of the plurality of first entries;

modifying the description of the target well log based at least partially upon the corpus of words, wherein the modified description of the target well log comprises one or more second n-grams;

identifying one or more second entries in the database having a measurement type assigned thereto that at least partially matches the modified description of the target well log, wherein identifying the one or more second entries comprises identifying the one or more first n-grams that match the one or more second n-grams;

assigning the measurement type of the one or more second entries to the target well log;

determining a search window in the target well log;

identifying one or more reference wells based at least partially upon the target well log;

predicting a plurality of locations of a target well marker in the target well log based at least partially upon the one or more reference wells, wherein the plurality of locations of the target well marker are within the search window;

determining similarity values for the plurality of locations of the target well marker;

automatically assigning the target well marker to at least one of the plurality of locations based at least partially upon the similarity values; and

automatically performing a wellsite activity based on the target well log, wherein the wellsite activity is at least one of:

modifying a trajectory of a wellbore;

drilling the wellbore in a subterranean formation based at least partially on the modified trajectory;

varying a weight on the drill bit (WOB);

varying a rate of penetration (ROP);

varying a type of fluid pumped into the wellbore;

varying a volumetric rate of fluid pumped into the wellbore; or

a combination thereof.

2 . The method of claim 1 , wherein predicting the plurality of locations of the target well marker comprises:

predicting a first location of the target well marker based at least partially upon a first reference well marker in a first of the one or more reference wells; and

predicting a second location of the target well marker based at least partially upon a second reference well marker in a second of the one or more reference wells.

3 . The method of claim 1 , further comprising generating or updating a model to include the assignment of the target well marker.

4 . A method, comprising:

determining a search window in a target well, wherein determining the search window in the target well comprises determining a first target well boundary and a second target well boundary;

identifying one or more reference wells based at least partially upon a location of the target well, wherein the one or more reference wells comprise a first reference well boundary, a second reference well boundary, and a reference well marker, wherein a predetermined confidence level exists that a reference well marker is positioned between the first reference well boundary and the second reference well boundary, and wherein a distance between the first and second reference well boundaries is less than a distance between the first target well boundary and the second target well boundary;

predicting a plurality of locations of a target well marker in the target well based at least partially upon the one or more reference wells, wherein the locations of the target well marker are within the search window, wherein the predetermined confidence level exists that the target well marker is positioned between the first target well boundary and the second target well boundary;

determining similarity values for the locations of the target well marker;

automatically assigning the target well marker to at least one of the plurality of locations based at least partially upon the similarity values; and

automatically controlling a wellsite activity based on the target well marker, wherein the wellsite activity is at least one of:

modifying a trajectory of a wellbore;

drilling the wellbore in a subterranean formation based at least partially on the modified trajectory;

varying a weight on the drill bit (WOB);

varying a rate of penetration (ROP);

varying a type of fluid pumped into the wellbore;

varying a volumetric rate of fluid pumped into the wellbore; or

a combination thereof.

5 . The method of claim 4 , further comprising:

determining a plurality of well signatures of the reference well marker in one of the one or more reference wells based at least partially upon a downhole measurement in the one or more reference wells; and

determining whether the well signatures are consistent with one another using a clustering algorithm.

6 . The method of claim 5 , wherein determining whether the plurality of well signatures are consistent with one another comprises determining that the plurality of well signatures are inconsistent with one another based at least partially upon the well signatures being present in a first cluster and a second cluster, the method further comprising determining that the first cluster is more accurate than the second cluster based at least partially upon the first cluster having a higher density of the well signatures than the second cluster.

7 . The method of claim 4 , wherein predicting the plurality of locations of the target well marker comprises:

predicting a first location of the target well marker based at least partially upon a first reference well marker in a first of the one or more reference wells; and

predicting a second location of the target well marker based at least partially upon a second reference well marker in a second of the one or more reference wells.

8 . The method of claim 4 , wherein predicting the plurality of locations of the target well marker comprises:

predicting a first location of the target well marker using a first algorithm based at least partially upon the reference well marker in the one or more reference wells; and

predicting a second location of the target well marker using a second algorithm based at least partially upon the reference well marker in the one or more reference wells, wherein the first algorithm and the second algorithm are different.

9 . The method of claim 4 , wherein the similarity values for the plurality of locations of the target well marker are determined based upon at least two of the one or more reference wells and at least two different algorithms.

10 . The method of claim 4 , wherein the target well marker is assigned to one of the plurality of locations having the highest similarity value.

11 . The method of claim 4 , further comprising generating or updating a model to include the assignment of the target well marker.

12 . The method of claim 4 , wherein the target well has a unit and a description assigned thereto, wherein the target well does not have a measurement type assigned thereto, and wherein the method further comprises:

identifying a plurality of first entries in a database, wherein the plurality of first entries have a unit and a measurement type assigned thereto, and wherein the units of the first entries are compatible with the unit of the target well;

generating a corpus of words from the measurement types of the first entries;

modifying the description of the target well based at least partially upon the corpus of words;

identifying one or more second entries in the database, wherein the one or more second entries have a measurement type assigned thereto, wherein the measurement type of the one or more second entries at least partially matches the modified description of the target well; and

assigning the measurement type of the one or more second entries to the target well.

13 . A method, comprising:

receiving a target well log having a unit and a description assigned thereto;

identifying a plurality of first entries in a database that have a unit and a measurement type assigned thereto, wherein the units of the plurality of first entries are compatible with the unit of the target well log;

generating a corpus of words based on the measurement types of the first entries, wherein generating the corpus of words comprises generating one or more n-grams based at least partially upon the measurement types of the first entries;

modifying the description of the target well log based at least partially upon the corpus of words;

identifying one or more second entries in the database having a measurement type assigned thereto that at least partially matches the modified description of the target well log;

automatically assigning the measurement type of the one or more second entries to the target well log; and

automatically performing a wellsite activity based on the target well log, wherein the wellsite activity is at least one of:

modifying a trajectory of a wellbore;

drilling the wellbore in a subterranean formation based at least partially on the modified trajectory;

varying a weight on the drill bit (WOB);

varying a rate of penetration (ROP);

varying a type of fluid pumped into the wellbore;

varying a volumetric rate of fluid pumped into the wellbore; or

a combination thereof.

14 . The method of claim 13 , wherein modifying the description of the target well log comprises removing one or more words from the description of the target well log that are not in the corpus of words.

15 . The method of claim 13 , wherein the one or more second entries comprise a subset of the plurality of first entries.

16 . The method of claim 13 , wherein the measurement type of the plurality of first entries comprises one or more first n-grams, wherein the modified description of the target well log comprises one or more second n-grams, and wherein identifying the one or more second entries comprises identifying the one or more first n-grams that match the one or more second n-grams.

17 . The method of claim 16 , wherein the one or more second entries comprise a plurality of second entries, and wherein assigning the measurement type comprises:

identifying a particular second entry whose measurement type has a highest number of the one or more first n-grams that match the one or more second n-grams; and

assigning the measurement type of the particular second entry to the target well log.

18 . The method of claim 13 , further comprising generating or updating a model to include the assigned measurement type.

19 . The method of claim 13 , further comprising:

determining a search window in the target well log;

identifying one or more reference wells based at least partially upon the target well log;

predicting a plurality of locations of a target well marker in the target well log based at least partially upon the one or more reference wells, wherein the plurality of locations of the target well marker are within the search window;

determining similarity values for the locations of the target well marker; and

assigning the target well marker to at least one of the locations based at least partially upon the similarity values.

Assignments (1)

Continuity (2)

References Cited (18)