Granted Patent B2

US 12,735,972 · App. 18/334,464 · Granted Sep 15, 2026

Inventors: Rafael Guedes de Carvalho (Katy, TX); Diego Fernando Patino Virano (Doha, QA)

Assignee: Schlumberger Technology Corporation

E21B44/00E21B2200/22

Abstract

Claims (48)

1 . A non-transitory computer readable medium having stored therein instructions executable by a computer system to cause the computer system to:

monitor a sensor suite, the sensor suite collecting drilling information about a drilling system during a wellsite operation, the drilling system performing drilling operations according to a drill plan, wherein the drilling information includes pressure measurements;

detect, via a machine learning model trained on knowledge in a knowledge database, a trigger event during the wellsite operation by identifying that at least a portion of the drilling information is outside of a threshold range of the drilling information, wherein the knowledge in the knowledge database includes explicit knowledge and tacit knowledge, and wherein the threshold range is a pressure threshold range and detecting the trigger event includes detecting that the pressure measurements are outside of the pressure threshold range;

in response to detecting the trigger event:

identify event data from the sensor suite at a time of the trigger event, the event data including unrelated data not relevant to the trigger event;

identify plan information from the drill plan that describes an activity being performed during the wellsite operation when the trigger event was detected; and

generate one of a plurality of alerts based on the event data, the plurality of alerts including a high pressure alert, a low pressure alert, and a trending pressure alert;

responsive to the one of the plurality of alerts, prompt a user associated with the one of the plurality of alerts to enter manual information about the trigger event, the manual information including new tacit knowledge not included in the drilling information or the event data;

prepare a recommendation to adjust drilling parameters based on the trigger event, wherein the recommendation includes adjusting a drilling fluid flow rate to return the pressure measurements to within the pressure threshold range;

implement, via a drilling integrator, the recommendation in the drilling system by automatically adjusting the drilling parameters at the drilling system and providing pump instructions to a pump to adjust the drilling fluid flow rate and return the pressure measurements to within the pressure threshold range;

add the drilling information, the event data, the plan information, the manual information, and the recommendation to the knowledge database; and

train the machine learning model on the drilling information, the event data, the plan information, and the manual information added to the knowledge database.

2 . The non-transitory computer readable medium of claim 1 , wherein detecting the trigger event includes detecting at least one of a drilling dysfunction, a likelihood of the drilling dysfunction, a near-miss event, or a deviation from the drill plan.

3 . The non-transitory computer readable medium of claim 1 , wherein prompting the user to enter the manual information includes prompting the user to enter at least one of a label of the trigger event, a categorization of the trigger event, or free text related to the trigger event.

4 . The non-transitory computer readable medium of claim 1 , wherein the event data comprises at least one of a time stamp, a current depth, measured drilling parameters, surface sensor readings, or an actual procedure being executed.

5 . The non-transitory computer readable medium of claim 1 , wherein the plan information comprises at least one of a planned activity, standard operating procedures, or planned drilling parameters.

6 . The non-transitory computer readable medium of claim 1 , wherein monitoring the sensor suite includes monitoring at least one of a multi-sensor system, manual entries by personnel in reporting systems, surface sensor information, downhole sensor information, recorded video, or the plan information.

7 . The non-transitory computer readable medium of claim 1 , wherein the instructions are executable by the computer system to cause the computer system to generate a likelihood of one or more trigger events occurring for a planned wellbore.

8 . The non-transitory computer readable medium of claim 1 , wherein the unrelated data includes weather and a crew on shift.

9 . The non-transitory computer readable medium of claim 1 , wherein the plurality of alerts include an alert type and an alert severity, and wherein the user associated with the one of the plurality of alerts has an associated level of management.

10 . A method for monitoring a drilling system, the method comprising:

performing a drilling operation;

collecting drilling information for the drilling operation from one or more sensors, wherein the drilling information includes pressure measurements;

applying the drilling information as input to a machine learning model trained on knowledge in a knowledge database to identify a trigger event, the trigger event including a deviation from a threshold range of at least a portion of the drilling information, wherein the knowledge in the knowledge database includes explicit knowledge and tacit knowledge, and wherein the threshold range is a pressure threshold range and identifying the trigger event includes detecting that the pressure measurements are outside of the pressure threshold range;

associating the trigger event with a knowledge type;

using the knowledge type, identifying event data from the drilling information at a time of the trigger event, the event data including unrelated data not relevant to the trigger event;

generating one of a plurality of alerts based on the event data, the plurality of alerts including a high pressure alert, a low pressure alert, and a trending pressure alert;

responsive to the one of the plurality of alerts, prompting a user associated with the one of the plurality of alerts to enter manual information about the trigger event, the manual information including new tacit knowledge not included in the drilling information or the event data;

using the knowledge type, preparing a recommendation to adjust drilling parameters based on the trigger event, wherein the recommendation includes adjusting a drilling fluid flow rate to return the pressure measurements to within the pressure threshold range;

adjusting the drilling parameters to return the drilling information to within the threshold range by adjusting the drilling fluid flow rate and returning the pressure measurements to within the pressure threshold range;

adding the drilling information, the event data, the manual information, and the recommendation to the knowledge database; and

training the machine learning model on the drilling information, the event data, and the manual information added to the knowledge database.

11 . The method of claim 10 , wherein adjusting the drilling parameters includes instructing a drilling operator to adjust a procedure.

12 . The method of claim 10 , wherein the one of the plurality of alerts includes a severity level based on at least one of the knowledge type or the deviation from the threshold range.

13 . A system comprising:

a drilling system that performs a drilling operation;

one or more sensors configured to sense drilling information about the drilling system;

a processor and memory, the memory including instructions which cause the processor to:

collect the drilling information for the drilling operation from the one or more sensors, wherein the drilling information includes pressure measurements;

apply the drilling information as input to a machine learning model trained on knowledge in a knowledge database to identify a trigger event, the trigger event including a deviation from a threshold range of at least a portion of the drilling information, wherein the knowledge in the knowledge database includes explicit knowledge and tacit knowledge, and wherein the threshold range is a pressure threshold range and identifying the trigger event includes detecting that the pressure measurements are outside of the pressure threshold range;

associate the trigger event with a knowledge type;

using the knowledge type, identify event data from the drilling information at a time of the trigger event, the event data including unrelated data not relevant to the trigger event;

generate one of a plurality of alerts based on the event data, the plurality of alerts including a high pressure alert, a low pressure alert, and a trending pressure alert;

responsive to the one of the plurality of alerts, prompt a user associated with the one of the plurality of alerts to enter manual information about the trigger event, the manual information including new tacit knowledge not included in the drilling information or the event data;

using the knowledge type, prepare a recommendation to adjust drilling parameters based on the trigger event, wherein the recommendation includes adjusting a drilling fluid flow rate to return the pressure measurements to within the pressure threshold range;

based on the knowledge type, adjust the drilling parameters to return the drilling information to within the threshold range by adjusting the drilling fluid flow rate and returning the pressure measurements to within the pressure threshold range;

add the drilling information, the event data, the manual information, and the recommendation to the knowledge database; and

train the machine learning model on the drilling information, the event data, and the manual information added to the knowledge database.

Assignments (1)

Continuity (2)

References Cited (21)